Download your result as a plain text file. Edit any field above and press Calculate to refresh it.

To add days to a date, count forward from the start date one day at a time and roll into the next month whenever you pass its last day.

Rule of thumb: 30 days is not one month. Add days when your document says days, and add months when it says months, because 31 January plus one month is 28 February while 31 January plus 30 days is 2 March.

Day counting has three traps. The first is whether the start day counts as day one, which shifts every answer by a day. The second is whether the days are calendar days or working days, since 90 working days is about 126 calendar days. The third is the end of the month, where adding a month to the 31st has to land somewhere real. This calculator handles all three and tells you which rule it applied.

What is 30 days from a date?

Thirty days from a date is that date plus 30 calendar days, weekends included. From 30 August 2026 it is Tuesday 29 September 2026. Thirty days is not one month: months run 28 to 31 days, so the two answers differ in eight months out of twelve.

How do you add days to a date manually?

Count the days left in the current month, subtract that from your total, then carry the remainder into the next month. From 30 August, one day reaches 31 August and the other 29 land on 29 September. Repeat month by month until the total is used up.

How many working days is 90 calendar days?

Ninety calendar days contains 64 or 65 working days once weekends are removed, depending on which weekday you start from. Read the other way, 90 working days takes about 126 calendar days. That gap of five weeks is why a contract that says days rather than working days is worth reading twice.

Does adding 30 days always change the weekday?

Yes, by exactly two days. Thirty divided by seven leaves a remainder of two, so a Sunday start becomes a Tuesday. Only multiples of seven leave the weekday untouched, which makes that the quickest free check on any date calculation you do.

What is the formula to add days to a date?

The formula is result date equals start date plus n days. Spreadsheets store a date as a day number, so the arithmetic is ordinary addition: in Excel it is =A2+30, in R as.Date("2026-08-30") + 30, and in Python date(2026, 8, 30) + timedelta(days=30).

How do I count 90 days from today?

Enter today's date as the start date, type 90 in the days field and press Calculate. The tool returns the exact date, the weekday it falls on and the ISO week number. Switch to business day mode first if your deadline counts working days only.

When you add days to a date you move forward along the calendar by a fixed number of whole days. The date is not a quantity you can average or split. It is a position on a line, and adding days slides you along that line. Every day you pass counts once, whatever month it belongs to, so the answer rolls into the next month automatically when you run past the last day of the current one.

Computers make this easy by turning a date into a plain number. The international standard for writing and ordering dates is ISO 8601, which is why every date field on this page uses the YYYY-MM-DD form. Excel counts days from 1 January 1900, so 30 August 2026 is the number 46264 and adding 30 gives 46294, which is 29 September 2026. Unix systems count seconds from 1 January 1970. R counts days from 1 January 1970. In every case the addition is ordinary arithmetic on a counter, and the calendar formatting is applied afterwards.

A worked one-liner: 30 August 2026 plus 30 days. August has 31 days, so 1 day takes you to 31 August, and the remaining 29 days take you to 29 September 2026, which is a Tuesday.

A note on the method this tool uses, and why. Three counting conventions are in common use and they disagree by a day or more, so a calculator has to pick one and say which. This one counts exclusively by default, meaning day 1 is the day after the start date, because that is what spreadsheets, programming languages and ordinary speech all do. It clamps month ends rather than rolling them forward, matching Excel's EDATE function and the Python datetime library. And it never moves a result off a weekend by itself, because whether a deadline rolls forward or backward is a term of your contract, not a property of the calendar. Tools that roll dates silently are the reason two parties can run the same sum and get different answers.

Who needs this and why. Lawyers count notice periods and limitation dates. Accountants count payment terms such as net 30 and net 90. Shipping teams promise delivery in working days. Clinics book follow-up appointments 6 weeks out. Students count backwards from a submission deadline. Project managers convert a plan measured in working days into calendar dates a client can read.

Days, weeks, months and years are not interchangeable. Days and weeks are fixed lengths, so they behave the same everywhere on the calendar. Months and years are variable, so adding them is a different operation with its own end of month rule. The table below is the whole difference in one place.

Conclusion: read your document carefully. The word it uses decides which row of this table you are in, and the rows genuinely disagree.

4.1 What your result date actually is

Your result date is the calendar day you reach after counting the number of days you asked for, starting from the day after your start date unless you switched on inclusive counting. It is a position, not a duration. The calculator also reports the weekday, because in practice the weekday is what decides whether a deadline is workable at all. A deadline that lands on a Sunday is a deadline that in effect lands on the Friday before or the Monday after, depending on the rule your document uses.

4.2 The weekday matters more than people expect

Adding a multiple of 7 days never changes the weekday, which is a useful check: if you add 28 days to a Tuesday and do not land on a Tuesday, something is wrong. Adding any other number moves the weekday by that number modulo 7. Adding 30 days moves you forward by two weekdays, because 30 divided by 7 leaves a remainder of 2. Adding 365 days in a common year moves you forward by one weekday, which is why your birthday shifts a day each year and two days across a leap year.

4.3 Calendar days versus working days

In calendar day mode every day counts, including Saturdays, Sundays and public holidays. In business day mode the calculator steps forward one day at a time and only counts a day when it is not a weekend day and not on your holiday list. The gap between the two grows quickly. Ten working days is 14 calendar days. Thirty working days is about 42. Ninety working days is about 126, which is more than four months rather than the three months people expect.

4.4 Why the start day rule changes the answer

By default day 1 is the day after your start date, which is how most spreadsheets, programming languages and everyday speech work: "in three days" said on Monday means Thursday. Many legal and contractual texts instead count the start day itself as day 1, which pulls every answer one day earlier. The calculator states which rule it used in the results table. If your document says "within 30 days of receipt", check whether receipt day counts, because a one day error on a filing deadline is not a rounding issue.

4.5 What the end of month clamp did to your answer

When you add months or years, the day of the month is kept if it exists in the target month and clamped to the last day if it does not. The arithmetic behind this was formalised by Dershowitz and Reingold (1990), whose calendrical calculations paper is the reference most date libraries still trace back to. 31 January plus one month gives 28 February, or 29 February in a leap year. This is what Excel's EDATE, Python's relativedelta and most database engines do, so the calculator matches them deliberately. One consequence surprises people: the operation is not reversible. Adding a month to 31 January gives 28 February, and subtracting a month from 28 February gives 28 January, not 31 January.

4.6 How the day of year and week number help

Fliegel and Van Flandern (1968) published the compact integer algorithm that converts a calendar date to a day number and back, which is why this arithmetic is fast enough to run as you type. The day of year runs from 1 to 365, or 366 in a leap year, and it is the fastest way to compare two dates in the same year without doing any calendar arithmetic. The ISO 8601 week number counts weeks starting on Monday, where week 1 is the week holding the first Thursday of the year. Payroll runs, retail plans and manufacturing schedules are very often expressed in ISO weeks, so having the number beside your date saves a lookup.

4.7 What to do next with your result

Three practical moves. First, check whether the result lands on a weekend or a holiday, and if it does, decide with the other party whether it rolls forward or backward before it becomes a dispute. Second, work backwards: if the result is a deadline, subtract your own preparation time in business day mode to get the date you actually need to start. Third, put the date in your calendar with the rule you used written in the note, because in three months you will not remember whether you counted the start day.

4.8 Common mistakes this calculator prevents

• Treating 30 days as one month, which is wrong in 8 months out of 12.
• Adding 12 months of 30 days and getting 360, then wondering why the year is short by five days.
• Counting working days by dividing calendar days by 5 and multiplying by 7, which ignores where the weekend actually falls.
• Forgetting 29 February, which is the single most common cause of an off by one date error in a spreadsheet.
• Pasting a date as text so the spreadsheet adds nothing at all and silently returns an error.
• Mixing day-month-year and month-day-year formats, so 03/04/2026 becomes two different days.

4.9 What this result does not tell you

It does not tell you whether the deadline is legally valid, whether a court would move it, or whether the other party counts the same way. It does not know about local public holidays unless you paste them in, and it does not know about half days, company shutdowns or time zones. A date has no time attached here, so an appointment at 23:00 on the result date is the same date as one at 09:00.

4.10 Practical judgement

Calculate the date, then read it back in words: "thirty days from Sunday 30 August 2026 is Tuesday 29 September 2026". Saying it out loud catches most mistakes immediately, because a wrong weekday sounds wrong. Then confirm which counting rule your document uses before you act on the number, and keep a note of the rule with the date.

▶ Run the calculator above to auto-fill all five templates with your own dates.

The worked example carried through all ten steps is: start on 30 August 2026 and add 30 days, which gives Tuesday 29 September 2026.

1. Enter your start date. The field is pre-filled with 30 August 2026 so the page is never blank. Click the calendar icon or type the date directly.
2. Choose add or subtract. Leave it on Add to move forward. Choose Subtract to work backwards from a deadline, which is the same arithmetic in the other direction.
3. Type the number of days. Enter 30 for the worked example. Whole numbers only, because a date is a whole day and half a day has no calendar meaning.
4. Add weeks, months or years if you need them. The four fields stack, so 1 month and 10 days is a single calculation. Leave them at zero for a plain day count.
5. Pick calendar days or business days. Calendar days count every day, which is what the worked example uses. Business days skip weekends, which is what shipping and many contracts mean.
6. Set the weekend rule if you chose business days. Saturday and Sunday is the default. Friday and Saturday suits much of the Middle East, and Sunday only suits some six day working weeks.
7. Paste any holidays into Advanced options, one date per line in YYYY-MM-DD form. They are skipped in business day mode. Weekend holidays are ignored, since that day is already skipped.
8. Decide whether the start day counts. The default is no, so 30 August plus 1 day is 31 August. Switch it on if your document counts the start day as day 1, which shifts the result one day earlier.
9. Press Calculate Date. The page opens with this worked example already answered, and any change you make clears the results so you never read stale numbers. Press Calculate and the result panel shows 29 September 2026 with its weekday, while the calculation strip shows your own numbers inside the formula.
10. Read the results and export. Section 2 gives the full breakdown, the weekday table and both charts. Download TXT gives the short answer, and Download Doc and Download PDF give the whole report.

Excel stores every date as a whole number counting days from 1 January 1900, so adding days to a date in Excel is ordinary addition. The date 30 August 2026 is stored as the number 46264, and adding 30 gives 46294, which Excel displays as 29 September 2026 once the cell is formatted as a date. Everything else in this section is built on that one idea.

8.1 The functions you need

8.2 Step by step

8.3 The awkward cases

Adding a month to the 31st. =EDATE(A2,1) on 31 January returns 28 February, and on 31 March returns 30 April. Excel never invents a 31 February. What surprises people is that the operation is not reversible: =EDATE(EDATE(A2,1),-1) starting from 31 January gives 28 January, not 31 January.

Adding years across 29 February. =EDATE("2028-02-29",12) gives 28 February 2029, because 2029 is not a leap year. =DATE(YEAR(A2)+1,MONTH(A2),DAY(A2)) gives 1 March 2029 instead, because DATE rolls an impossible day forward rather than clamping it. Both are defensible, and they disagree. Pick one and say which.

Negative and zero arguments. =WORKDAY(A2,0) returns the start date itself even when the start date is a Saturday, which is not what most people expect. =WORKDAY(A2,-10) counts backwards correctly. If you need the next working day on or after a date, use =WORKDAY(A2-1,1).

Blank cells. An empty A2 is read as zero, which Excel formats as 0 January 1900. Guard the formula with =IF(A2="","",A2+30) so a half filled sheet does not display nonsense dates.

Dates before 1900. Excel cannot store them as dates at all in the 1900 system, so historical work has to use text or a different tool. This calculator handles year 1 onwards, so the two will disagree on anything before 1900.

8.4 Why your Excel answer might differ from R or Python

Note the JavaScript row: setMonth does not clamp, so 31 January plus one month gives 3 March, not 28 February. That is the single biggest disagreement in this table.

8.5 Charting dates in Excel

1. Put your milestone names in one column and their calculated dates in the next.
2. Select both columns, then Insert → Charts → Bar for a simple timeline, or a Scatter chart if you want the dates spaced by their real gaps.
3. Right click the date axis, choose Format Axis, and set the bounds as dates rather than numbers, otherwise Excel shows serial numbers on the axis.
4. Set the axis number format to dd mmm so the labels stay readable when the chart is narrow.
5. For a Gantt style bar, plot the start date as a series with no fill and the duration in days as the visible bar on top of it.

Chart 2 in Section 2 of this page already draws the common offsets from your start date, which is usually quicker than rebuilding it in Excel.

8.6 Excel errors you will probably hit

Google Sheets: every formula in this section works unchanged, including WORKDAY, WORKDAY.INTL, EDATE, EOMONTH and ISOWEEKNUM. The only real difference is that Sheets counts from 30 December 1899 and has no 1900 leap year bug, so the two disagree only on dates before March 1900.

In R a Date object is stored as a number of days since 1 January 1970, so adding days to a date in R is plain addition: as.Date("2026-08-30") + 30. Everything harder than that, meaning calendar months, business days and the end of month clamp, needs a few extra lines, and they are all in the script below. It uses base R only, so there is nothing to install, and the behaviour of every function is documented in the base R date reference.

9.1 The complete script

R, base only, saves add_days_plot.png

# ============================================================
# Add days to a date, complete worked example
# Base R only. Prints a plain-English verdict and saves one
# 300 dpi figure called add_days_plot.png
# ============================================================

# ---- 1. YOUR DATA: change only these five lines -------------
start_date    <- as.Date("2026-08-30")   # start date
days_to_add   <- 30                      # whole days, negative to subtract
months_to_add <- 0                       # calendar months, clamped
business_only <- FALSE                   # TRUE skips weekends and holidays
holidays      <- as.Date(c("2026-09-07"))
# -------------------------------------------------------------

# Alternative: read a column of start dates from a CSV
# dat <- read.csv("dates.csv", stringsAsFactors = FALSE)
# dat <- dat[!is.na(dat$start), ]          # drop missing values
# start_date <- as.Date(dat$start[1])

add_months <- function(d, k) {
  lt  <- as.POSIXlt(d)
  tot <- lt$mon + k
  y   <- lt$year + 1900 + tot %/% 12
  m   <- tot %% 12 + 1
  last <- as.integer(format(seq(as.Date(sprintf("%04d-%02d-01", y, m)),
                                by = "month", length.out = 2)[2] - 1, "%d"))
  as.Date(sprintf("%04d-%02d-%02d", y, m, min(lt$mday, last)))
}

is_workday <- function(d) {
  wd <- as.integer(format(d, "%u"))        # 1 = Monday ... 7 = Sunday
  wd <= 5 & !(d %in% holidays)
}

stepped <- add_months(start_date, months_to_add)
clamped <- as.integer(format(stepped, "%d")) != as.integer(format(start_date, "%d"))

if (business_only) {
  step <- if (days_to_add >= 0) 1 else -1
  d <- stepped; counted <- 0
  while (counted < abs(days_to_add)) {
    d <- d + step
    if (is_workday(d)) counted <- counted + 1
  }
  result <- d
} else {
  result <- stepped + days_to_add
}

span     <- as.integer(result - start_date)
all_days <- seq(start_date + 1, start_date + abs(span), by = "day")
if (span < 0) all_days <- seq(start_date - abs(span), start_date - 1, by = "day")
wd       <- as.integer(format(all_days, "%u"))
weekend  <- sum(wd > 5)
hol_hits <- sum(wd <= 5 & all_days %in% holidays)
working  <- length(all_days) - weekend - hol_hits

cat("Start date      :", format(start_date, "%A %d %B %Y"), "\n")
cat("Added           :", months_to_add, "months and", days_to_add,
    if (business_only) "business days" else "calendar days", "\n")
cat("Result date     :", format(result, "%A %d %B %Y"), "\n")
cat("Calendar span   :", span, "days\n")
cat("Working days    :", working, "\n")
cat("Weekend days    :", weekend, "\n")
cat("Holidays skipped:", hol_hits, "\n")
cat("Day of year     :", as.integer(format(result, "%j")), "\n")
cat("ISO week number :", as.integer(format(result, "%V")), "\n")
if (clamped) cat("NOTE: the day of the month was clamped to the last day of the target month.\n")
cat("\nVERDICT: the result date is", abs(span), "calendar days",
    if (span >= 0) "after" else "before",
    "the start date, and it lands on a", format(result, "%A"), "\n")

# ---- 2. FIGURE: one 300 dpi PNG -----------------------------
offsets <- c(7, 14, 30, 45, 60, 90, 180, 365)
dates   <- start_date + offsets
labs    <- format(dates, "%d %b %Y")
cols    <- ifelse(offsets == days_to_add, "#16a34a", "#93c5fd")

png("add_days_plot.png", width = 9, height = 5, units = "in", res = 300)
par(mar = c(4.5, 6, 3.5, 2))
bp <- barplot(rev(offsets), horiz = TRUE, col = rev(cols), border = "#1e3a8a",
              names.arg = rev(paste(offsets, "days")), las = 1, xlim = c(0, 430),
              xlab = paste("Days added to", format(start_date, "%d %B %Y")),
              main = "Where common day offsets land", col.main = "#1e3a8a")
text(rev(offsets) + 8, bp, rev(labs), adj = 0, cex = 0.85, col = "#1f2937")
dev.off()
cat("\nSaved add_days_plot.png at 300 dpi\n")

9.2 What each part of the script does

9.3 What the figure shows

The script writes add_days_plot.png to your working directory. It is a horizontal bar chart of the common day offsets, 7 through 365, with the resulting calendar date printed at the end of each bar. The bar matching the number of days you actually asked for is drawn in green so it stands out from the rest.

Read it as a sanity check rather than as a result. If your requested offset lands somewhere that looks wrong relative to its neighbours, you have almost certainly typed the wrong number of days.

9.4 Problems and fixes

9.5 One-liners worth remembering

Adding days to a date in Python is a one-liner with timedelta: date(2026, 8, 30) + timedelta(days=30). What timedelta deliberately does not have is a month unit, because a month has no fixed length. The rule is set out in the timedelta documentation. The script below adds the month arithmetic, the end of month clamp and a business day path, and it produces exactly the same numbers and the same figure as the R script above.

10.1 The complete script

Python 3, saves add_days_plot.png

# ============================================================
# Add days to a date, complete worked example
# Prints a plain-English verdict and saves one 300 dpi figure
# ============================================================
from datetime import date, timedelta
import calendar
import numpy as np
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt

# ---- 1. YOUR DATA: change only these five lines -------------
start_date = date(2026, 8, 30)   # start date, year, month, day
days_to_add = 30                 # whole days to add, negative to subtract
months_to_add = 0                # calendar months, clamped at end of month
business_only = False            # True skips weekends and holidays
holidays = ["2026-09-07"]        # ISO dates to skip in business mode
# -------------------------------------------------------------

# Alternative: read a column of start dates from a CSV
# import pandas as pd
# df = pd.read_csv("dates.csv", parse_dates=["start"])
# df = df.dropna(subset=["start"])          # drop missing values
# start_date = df["start"].iloc[0].date()

hol = np.array(holidays, dtype="datetime64[D]")

def add_months(d, k):
    """Add k calendar months, clamping the day to the last valid day."""
    total = d.month - 1 + k
    y = d.year + total // 12
    m = total % 12 + 1
    return date(y, m, min(d.day, calendar.monthrange(y, m)[1]))

stepped = add_months(start_date, months_to_add)
clamped = (stepped.day != start_date.day)

if business_only:
    result = np.busday_offset(np.datetime64(stepped, "D"), days_to_add,
                              roll="forward", holidays=hol).astype("datetime64[D]").astype(date)
else:
    result = stepped + timedelta(days=days_to_add)

span = (result - start_date).days
all_days = [start_date + timedelta(days=i) for i in range(1, abs(span) + 1)]
weekend = sum(1 for d in all_days if d.weekday() >= 5)
hol_hits = sum(1 for d in all_days if d.weekday() < 5 and str(d) in holidays)
working = len(all_days) - weekend - hol_hits

print("Start date      :", start_date.strftime("%A %d %B %Y"))
print("Added           :", months_to_add, "months and", days_to_add,
      "business days" if business_only else "calendar days")
print("Result date     :", result.strftime("%A %d %B %Y"))
print("Calendar span   :", span, "days")
print("Working days    :", working)
print("Weekend days    :", weekend)
print("Holidays skipped:", hol_hits)
print("Day of year     :", result.timetuple().tm_yday)
print("ISO week number :", result.isocalendar()[1])
if clamped:
    print("NOTE: the day of the month was clamped to the last day of the target month.")
print()
if span > 0:
    print("VERDICT: the result date is", span, "calendar days after the start date,",
          "and it lands on a", result.strftime("%A") + ".")
else:
    print("VERDICT: the result date is", abs(span), "calendar days before the start date.")

# ---- 2. FIGURE: one 300 dpi PNG -----------------------------
offsets = [7, 14, 30, 45, 60, 90, 180, 365]
dates = [start_date + timedelta(days=o) for o in offsets]
labels = [d.strftime("%d %b %Y") for d in dates]
colors = ["#16a34a" if o == days_to_add else "#93c5fd" for o in offsets]

fig, ax = plt.subplots(figsize=(9, 5))
bars = ax.barh([str(o) + " days" for o in offsets], offsets, color=colors,
               edgecolor="#1e3a8a", linewidth=0.8)
for bar, lab in zip(bars, labels):
    ax.text(bar.get_width() + 5, bar.get_y() + bar.get_height() / 2, lab,
            va="center", fontsize=9, color="#1f2937")
ax.set_xlabel("Days added to " + start_date.strftime("%d %B %Y"))
ax.set_title("Where common day offsets land", fontsize=13, color="#1e3a8a")
ax.set_xlim(0, 430)
ax.invert_yaxis()
ax.spines[["top", "right"]].set_visible(False)
fig.tight_layout()
fig.savefig("add_days_plot.png", dpi=300)
print("\nSaved add_days_plot.png at 300 dpi")

10.2 What each part of the script does

10.3 What the figure shows

The chart written to add_days_plot.png is identical in content to the R figure shown in section 9.3: one horizontal bar per common offset, the resulting date printed at the end of the bar, and your own offset drawn in green. Run both scripts on the same inputs and the two PNGs will carry the same dates.

10.4 Problems and fixes

10.5 One-liners worth remembering

11.1 Add days to a date: a ready reckoner from 30 August 2026

Conclusion: read the weekday column before you commit to a date. Five of these results land on a weekend, and a weekend deadline is usually a Monday deadline in practice.

11.2 Days in each month and the running day of year

Conclusion: only July and August are two 31 day months in a row, which is why a 62 day span is possible only across that pair. Everywhere else 30 days and 31 days alternate.

11.3 Adding 30 days against adding one calendar month, every month of 2026

Conclusion: starting on the 15th, the two answers match only when the following month has exactly 30 days. In every 31 day month they differ by one day, and in February they differ by two or three.

11.4 Business days converted to calendar days, from 30 August 2026

Conclusion: the rough estimate is close but almost never exact, because it cannot know where the weekends fall. Use it for planning and the real count for anything contractual.

11.5 Common payment and notice terms, counted from 30 August 2026

Conclusion: the two right hand columns are the reason terms get disputed. Net 90 counted as calendar days is 28 November 2026, and counted as working days it is 1 January 2027, five weeks later.

11.6 How far the weekday moves when you add days

Conclusion: if you add a multiple of 7 and the weekday changes, you have made an arithmetic error. This is the cheapest check available on any date calculation.

11.7 Leap years, 2020 to 2060

Conclusion: the last three rows are the century rule. 1900 and 2100 are not leap years because they do not divide by 400, and Excel is wrong about 1900 by design, for backward compatibility with an old spreadsheet.

A date calculation is only as good as the two things you feed it: the start date and the counting rule. Both are read off a document, and both are read wrongly more often than people expect.

1. Find the start date in the document, not in your memory. An invoice date, a service date and a receipt date are three different days, and contracts pick one of them deliberately.
2. Check the date order before you type it. 03/04/2026 is 3 April in the UK, India and most of Europe, and 4 March in the United States. If the source is ambiguous, find another date in the same document with a day above 12, which settles the order immediately.
3. Convert the date to ISO form as you go. Writing 2026-08-30 in your working notes removes the ambiguity permanently and matches what every tool on this page expects.
4. Read the counting word. Days, working days, business days, clear days and calendar days are five different phrases and at least three different rules. Copy the exact phrase into your notes.
5. Decide whether the start day counts. Look for the words from, after, within and inclusive. From and inclusive usually mean the start day counts. After usually means it does not.
6. Get the right holiday list. Public holidays differ by country and often by region within a country. Use the list for the place where the obligation is performed, not where you are sitting.
7. Check whether a weekend deadline rolls. Many rules move a deadline landing on a weekend or holiday to the next working day; some move it backwards instead. This calculator does not roll automatically, because the rule belongs to your document, not to the arithmetic.
8. Watch the time zone if a time is attached. A deadline of midnight in one place is the previous afternoon in another. Dates alone have no time zone, which is usually a simplification in your favour, but not always.
9. Write down which rule you used next to the answer. In three months you will remember the date and not the rule, and the rule is what makes the date defensible.
10. Check the answer backwards. Subtract the same number of days from your result with the same settings. If you do not land back on the start date exactly, one of your settings changed between the two runs.

One input per field, whole numbers only, and dates in YYYY-MM-DD form in the holiday box.

Common mistakes when gathering the inputs:

• Using today's date as the start date when the document names a different one.
• Copying a date from a spreadsheet cell that was displaying a different format from the value it held.
• Assuming the other party uses the same public holiday calendar.
• Counting a half day shutdown as a full holiday, or the reverse.
• Typing 12.5 days because the plan said two and a half weeks, when what you needed was 18 days.
• Mixing months and days in the same clause, such as three months and 15 days, without saying which is applied first.

Use this add days to date calculator when:

• ✓ You need the exact date a fixed number of days after or before another date.
• ✓ The number is written in days, weeks, months or years, or a mixture of them.
• ✓ You need to skip weekends, and possibly a list of public holidays.
• ✓ You need the weekday, the day of the year or the ISO week number of the answer.
• ✓ You are working backwards from a deadline to find a start date.
• ✓ You want the same calculation written out for Excel, R or Python so you can repeat it in bulk.

Do not use it when:

• ✗ You need the number of days between two dates you already have. That is the reverse calculation.
• ✗ Times of day matter, for example a deadline at 17:00 across time zones.
• ✗ You need a non-Gregorian calendar such as the Hijri, Hebrew or Japanese imperial calendar.
• ✗ You need the deadline to roll automatically when it lands on a weekend, because the rolling rule belongs to your document.
• ✗ You need legal certainty on a limitation date. Use this to prepare, then have the date confirmed.
• ✗ Dates before 1583, where the change from the Julian to the Gregorian calendar makes any single answer contestable.

Four worked situations:

1. Invoice terms. Net 30 from an invoice dated 30 August 2026 is 29 September 2026 in calendar days. Use calendar day mode.
2. Shipping promise. Ten working days from the same date is 11 September 2026. Use business day mode with your own holiday list.
3. Medical follow up. Six weeks after an appointment keeps the same weekday, because 42 divides by 7. Use the weeks field.
4. Tenancy notice. Two months from 31 January 2026 is 31 March 2026, because March is long enough to keep the 31st. Use the months field, not 60 days.

Assumptions

1. The proleptic Gregorian calendar applies throughout. If it fails, dates before 1583 shift by up to 13 days, because the calendar reform was adopted in different countries in different years.
2. Day 1 is the day after the start date unless you say otherwise. If it fails, every result is one day late.
3. Months are added with an end of month clamp. If your document expects a roll forward instead, results from the 29th, 30th and 31st are one to three days early.
4. Years and months are applied before weeks and days. If your document expects the other order, mixed offsets can differ by a day or two around month ends.
5. The weekend is the two days you selected, every week. If it fails, for example where a working week varies by season, the business day count drifts.
6. The holiday list you pasted is complete and correct. If it fails, business day results are early by one day for each missing holiday.
7. Dates carry no time of day. If it fails, a deadline expressed as a time in another time zone can fall on a different date.

Limitations

• It does not roll a weekend or holiday result to the next working day, because that rule belongs to your contract. Read the result, then apply your own rule. A date difference calculator will confirm the new gap, and the StatsUnlock calculator library is where those sibling tools are published as they go live.
• It has no built-in holiday calendar. Public holidays vary by country and region and change from year to year, so they have to be pasted in. An official government holiday list is the right source.
• It works only in the Gregorian calendar. For Hijri, Hebrew, Persian or Japanese imperial dates you need a calendar conversion tool.
• It handles whole days only. For hours and minutes across time zones you need a time zone calculator.
• It gives no legal advice. Limitation periods, court deadlines and statutory notice periods have their own counting rules that can differ from every convention on this page.
• It does not compare two dates. To find the gap between dates you already have, use a date difference calculator rather than working backwards from this one.
• Very large offsets are capped at year 9999, which is a limit of the date representation rather than of the arithmetic.

To add days to a date is the simplest calculation on this whole site, and it is still the one that produces the most disputes. The arithmetic never fails. What fails is the reading of the question: whether the start day counts, whether the days are calendar days or working days, and whether the document meant days at all when it wrote a number that looks like a month.

The habit worth building is to say the answer out loud with its weekday attached. Thirty days from Sunday 30 August 2026 is Tuesday 29 September 2026. A wrong weekday sounds wrong immediately, which is why naming it catches errors that staring at the number never will. Adding a multiple of seven should never change the weekday at all, and that free check has saved more deadlines than any amount of double entry.

The second habit is to record the rule beside the date. A date on its own is an assertion. A date with the words thirty calendar days, start day not counted, weekend Saturday and Sunday beside it is a calculation anyone can check and nobody can argue with. It takes one extra line in an email and it removes an entire category of argument later.

Where a result lands on a weekend or a public holiday, stop and decide what happens next before you rely on the date. This calculator deliberately does not move it for you, because moving it forward and moving it backward are both real conventions and only your document knows which one applies. Getting that agreed in advance costs a two line email; getting it wrong costs a missed deadline.

Use the calculator at the top of this page whenever you need to add days to a date, switch to business day mode when weekends should not count, and copy the Excel, R or Python version from Sections 8 to 10 when you need to run the same calculation over a whole column of dates rather than one.

Twenty peer-reviewed journal articles on calendars, date representation, temporal reasoning and date handling in software. Standards documents and official holiday calendars are named in Section 11 and Section 13 rather than here.

1. Allen, J. F. (1983). Maintaining knowledge about temporal intervals. Communications of the ACM, 26(11), 832-843. https://doi.org/10.1145/182.358434
2. Allen, J. F. (1984). Towards a general theory of action and time. Artificial Intelligence, 23(2), 123-154. https://doi.org/10.1016/0004-3702(84)90008-0
3. Bettini, C., Jajodia, S., & Wang, X. S. (1998). A general framework for time granularity and its application to temporal reasoning. Annals of Mathematics and Artificial Intelligence, 22(1), 29-58. https://doi.org/10.1023/A:1018938900494
4. Dyreson, C. E., Evans, W. S., Lin, H., & Snodgrass, R. T. (2000). Efficiently supporting temporal granularities. IEEE Transactions on Knowledge and Data Engineering, 12(4), 568-587. https://doi.org/10.1109/69.868908
5. Fliegel, H. F., & Van Flandern, T. C. (1968). A machine algorithm for processing calendar dates. Communications of the ACM, 11(10), 657. https://doi.org/10.1145/364096.364097
6. Goralwalla, I. A., Leontiev, Y., Ozsu, M. T., Szafron, D., & Combi, C. (2001). Temporal granularity: Completing the puzzle. Journal of Intelligent Information Systems, 16(1), 41-63. https://doi.org/10.1023/A:1008737019055
7. Snodgrass, R. T., & Ahn, I. (1986). Temporal databases. Computer, 19(9), 35-42. https://doi.org/10.1109/MC.1986.1663327
8. Jensen, C. S., & Snodgrass, R. T. (1999). Temporal data management. IEEE Transactions on Knowledge and Data Engineering, 11(1), 36-44. https://doi.org/10.1109/69.755613
9. Ohlbach, H. J., & Gabbay, D. (1998). Calendar logic. Journal of Applied Non-Classical Logics, 8(4), 291-324. https://doi.org/10.1080/11663081.1998.10510949
10. Reingold, E. M., Dershowitz, N., & Clamen, S. M. (1993). Calendrical calculations, II: Three historical calendars. Software: Practice and Experience, 23(4), 383-404. https://doi.org/10.1002/spe.4380230404
11. Dershowitz, N., & Reingold, E. M. (1990). Calendrical calculations. Software: Practice and Experience, 20(9), 899-928. https://doi.org/10.1002/spe.4380200905
12. Richards, E. G. (2013). Calendars and their reform. Astronomy & Geophysics, 54(3), 3.28-3.30. https://doi.org/10.1093/astrogeo/att088
13. Terenziani, P., & Snodgrass, R. T. (2004). Reconciling point-based and interval-based semantics in temporal relational databases. IEEE Transactions on Knowledge and Data Engineering, 16(5), 540-551. https://doi.org/10.1109/TKDE.2004.1277816
14. Verhagen, M., Sauri, R., Caselli, T., & Pustejovsky, J. (2010). Temporal expression recognition and normalization. Language Resources and Evaluation, 47(1), 123-152. https://doi.org/10.1007/s10579-012-9198-8
15. Strotgen, J., & Gertz, M. (2013). Multilingual and cross-domain temporal tagging. Language Resources and Evaluation, 47(2), 269-298. https://doi.org/10.1007/s10579-012-9179-y
16. Chen, T.-Y., Chen, Y.-M., & Wang, C.-B. (2011). A formal virtual enterprise access control model based on temporal constraints. Expert Systems with Applications, 38(6), 7503-7515. https://doi.org/10.1016/j.eswa.2010.12.109
17. Ramaswamy, S. (1997). Efficient indexing for constraint and temporal databases. Lecture Notes in Computer Science, 1186, 419-431. https://doi.org/10.1007/3-540-62222-5_59
18. Ahmad, K., & Cheng, D. (2007). Calendar-based temporal reasoning for financial text mining. Expert Systems with Applications, 33(3), 707-718. https://doi.org/10.1016/j.eswa.2006.06.017
19. Wang, X. S., Bettini, C., Brodsky, A., & Jajodia, S. (1997). Logical design for temporal databases with multiple granularities. ACM Transactions on Database Systems, 22(2), 115-170. https://doi.org/10.1145/249978.249979
20. Combi, C., Franceschet, M., & Peron, A. (2004). Representing and reasoning about temporal granularities. Journal of Logic and Computation, 14(1), 51-77. https://doi.org/10.1093/logcom/14.1.51