There are 3,600 seconds in one hour, and that number is not a rounding or a convention you can opt out of — it falls straight out of how the hour itself is defined. One hour is 60 minutes, one minute is 60 seconds, and 60 × 60 = 3,600. That holds for the standard clock hour under Coordinated Universal Time (UTC), which is the time scale most of the world's computer systems, airlines, and financial markets synchronize to. The figure never drifts by a second or two the way the calendar year does. Once you understand why the hour and the second were built on the same base-60 foundation, the arithmetic stops being trivia and starts being useful context.
The 60-Minute Hour and 60-Second Minute Both Trace to Babylonian Sexagesimal Counting
The number 3,600 isn't arbitrary. It comes from a counting system built on 60, which historians trace to ancient Mesopotamian scribes and astronomers who used base-60 numeration. They likely favored it because 60 divides evenly by 1, 2, 3, 4, 5, 6, 10, 12, 15, 20, 30, and 60 — a convenience when splitting anything into fractions. That same divisor-heavy structure passed through Hellenistic astronomy into the Roman and later European world. The division of the hour into 60 minutes and the minute into 60 seconds is a descendant of that tradition, not a modern invention. When you ask how many seconds an hour contains, you're really asking a question about a framework that's over two millennia old.
The consistency matters here. Because both subdivisions use 60, the conversion is a clean single multiplication: 60 minutes × 60 seconds = 3,600. There's no awkward remainder, no carry, no lunar-style adjustment. That structural simplicity is why the 3,600 figure shows up everywhere from payroll software to physics problems without a correction factor. A calendar year needs a leap day because a year is roughly 365.2422 days, and a day needs leap seconds because the Earth's rotation wobbles. The standard hour carries no such baggage. It's an exact integer count within the minute system.
The number also anchors a whole family of conversions once it's committed to memory. A day is 24 × 3,600 = 86,400 seconds. A week is 604,800 seconds. A 40-hour workweek is 144,000 seconds. These aren't separate facts to memorize; they all cascade from the same 3,600-second unit. That's the practical payoff of understanding the origin rather than just trusting a lookup. If you know the hour is 3,600 seconds, you can rebuild the day total in your head in two seconds.
One caveat worth naming: the historical account above is a mainstream summary, not a precise scholarly citation of a specific tablet or date. If you need the primary sources, the Museum of the History of Science at Oxford and the mathematical-history literature cover the sexagesimal lineage in detail. The arithmetic itself needs no citation at all — it's definitional.
All Three Conversion Routes Return the Same Integer: A Worked Calculation
This is the section that's most useful to verify once and then write on a sticky note. There are three standard routes to the same answer, and all three land on 3,600 with no remainder. The table below lines them up side by side so you can see the method, the arithmetic, and the result in one place.
| Method | What you multiply | Arithmetic | Result |
|---|---|---|---|
| Minutes to seconds | 60 minutes per hour × 60 seconds per minute | 60 × 60 | 3,600 seconds |
| Seconds-per-minute repeated | Add 60 seconds sixty times | 60 + 60 + … (60 terms) | 3,600 seconds |
| Fraction of a day | 86,400 seconds per day ÷ 24 hours per day | 86,400 ÷ 24 | 3,600 seconds |
The minute route is the one most people reach for, and it's the cleanest. The repeated-addition route is the way the arithmetic actually works underneath the multiplication; it's not a different answer, just a slower way to get it. The day-fraction route is the one that trips people up because it requires knowing the 86,400-second day figure first.
Start with the simplest case. If you multiply 60 minutes by 60 seconds per minute, you're just stating the definition of both units and combining them. You can do this in a spreadsheet cell as =60*60, which returns 3,600. You can do it in a Python shell with 60 * 60, which returns 3600. You can do it mentally. There's no ambiguity and no rounding at any step. That's why the answer is reliable no matter which platform you compute it on.
The repeated-addition route is a useful sanity check when you don't trust your multiplication. Count sixty seconds to a minute, then do it sixty times. If you lose your place, you've found a human error, not a mathematical one. The result must be 3,600, and if it isn't, you miscounted. This is the kind of verification that doesn't require a calculator and doesn't require trusting a source, which is the strongest form of verification available for a question this basic.
The day-fraction route is worth knowing because it generalizes. If someone tells you there are 86,400 seconds in a day — a standard figure in computing and astronomy contexts — dividing by 24 gives you 3,600. That's useful because you might encounter the day total first, in a log file or a programming reference, and need to back into the hour figure. All three routes are arithmetic and all three agree, which is why the 3,600 figure isn't going to change based on which one you pick.
A Non-Leap Hour Is Exactly 3,600 SI Seconds; the Leap Second Is the Rare Exception
Here's where the count gets a wrinkle, and it's the wrinkle most basic references skip. A standard, non-leap hour is exactly 3,600 seconds under the International System of Units (SI), because the SI second is the base unit and the minute and hour are defined as 60 and 3,600 of them. That's the figure to use in nearly every practical context. But on extremely rare occasions, an hour can contain 3,599 or 3,601 seconds, because a device has to insert or remove a leap second to keep atomic time aligned with the Earth's rotation. If you're doing payroll, homework, or scheduling, ignore this entirely. If you're writing timestamp code that touches leap-second handling, don't.
The reason a leap second exists is that the Earth's rotation isn't perfectly uniform. The day is slowing slightly over geological timescales, and the mismatch between atomic time (TAI) and astronomical time (UT1) has to be reconciled. Historically, the International Earth Rotation and Reference Systems Service (IERS) announced additions of leap seconds. A leap second makes a specific UTC minute 61 seconds long, and therefore that hour 3,601 seconds. The number of leap seconds actually added over the past several decades is small — well under a hundred since the practice began — so the overwhelming majority of hours you'll ever encounter are exactly 3,600 seconds.
The future of the leap second is itself in flux. In 2022, the General Conference on Weights and Measures voted to phase out leap seconds by 2035, meaning the organizations responsible for timekeeping may stop inserting them in UTC and instead tolerate a growing offset between UTC and astronomical time. If that plan holds, then "every hour is exactly 3,600 seconds" becomes true at the level of regulation, not just in practice. Until then, the rare edge case stands: a single hour can be off by one second.
If you're testing this yourself, the practical verification path is the IERS Bulletin C, which publishes leap-second announcements, and any official time source that reports UTC. Consumer devices silently absorb leap seconds in their firmware and simply display the next second, so you won't notice a difference at 3,600 vs. 3,601 in daily life. The distinction only matters to precision-sensitive systems — GPS receivers, financial exchange timestamps, and scientific instruments. For everything else, 3,600 is the answer and the conclusion holds.
A Day Computes to 86,400 Seconds, and That Figure Anchors Most Hour Conversions
Most people who ask how many seconds are in an hour are actually building toward a bigger number — a day, a shift, a billing period. The 3,600-second hour is the building block, and the day is the most common next stop. A standard 24-hour day is 24 × 3,600 = 86,400 seconds, which is the figure that shows up in payroll timekeeping, computer logs, and astronomy references. That's not a separate fact to look up; it's a direct consequence of the hour count. If you know one, you know the other.
| Time span | Formula | Seconds |
|---|---|---|
| 1 hour | 3,600 × 1 | 3,600 |
| 8-hour workday | 3,600 × 8 | 28,800 |
| 12-hour shift | 3,600 × 12 | 43,200 |
| 24-hour day | 3,600 × 24 | 86,400 |
| 7-day week | 86,400 × 7 | 604,800 |
Start with an eight-hour workday. Eight times 3,600 is 28,800 seconds. A twelve-hour shift is 43,200 seconds. A 40-hour week is 144,000 seconds, and a standard calendar week is 604,800 seconds. These numbers are useful when you're estimating time budgets, calculating machine uptime, or converting a duration you see in one unit into another. The arithmetic is trivial once the 3,600 anchor is in place, which is why memorizing the hour figure pays off across the whole time system.
The day figure has its own edge cases, and they matter for the same reason the leap-second edge case does. A daylight saving time "spring forward" day is 23 hours, or 82,800 seconds. A "fall back" day is 25 hours, or 90,000 seconds. Those aren't leap seconds; they're civil time adjustments that most timekeeping systems represent as a change in the offset, not a change in the underlying second count. If you're computing elapsed time across a DST boundary, you need to account for the shift, or your total will be an hour off in either direction.
If you ever need the precise number for a day and you're not sure whether a leap second landed, the safe approach is to check the UTC date against IERS announcements rather than trusting a generic "86,400." For everyday purposes, 86,400 is the number and it's accurate to within a second on essentially every day you'll ever encounter. That reliability is why the day figure keeps circulating without a caveat attached, even though the caveat still technically exists.
Spreadsheets and Code Convert Reliably to 3,600; the Failures Come From How You Feed Them the Input
The conversion 1 hour = 3,600 seconds is trivial for a computer. What fails isn't the arithmetic; it's the way data gets into or out of the conversion. Spreadsheet timestamps, text-formatted durations, and time-zone-aware libraries each have specific quirks that can turn a clean 3,600 into something else. If you're writing formulas in Excel or Google Sheets, write them so they accept a plain hour as a number, and format the input cell as a decimal, not a time. A time-formatted cell holds a fraction of a day, not an integer hour count, so multiply-by-3,600 on a timestamp fails.
Start with the simplest spreadsheet formula. Type 3 in a cell, then in another cell write =A13600, and the result is 10,800. That's the clean path for a plain numeric count. If instead you type 3:00 into a cell and then use =A23600, you might get 0.125 because that colon notation is being read as a time value, not a decimal count of hours. The fix is to convert the time value back to decimal hours first: =HOUR(A2)+MINUTE(A2)/60 gives you the decimal hour count, and then multiply by 3,600. The failure isn't in the arithmetic at 3,600; it's in how the spreadsheet stored the input.
On the code side, the standard library approach depends on the language. In Python's datetime module, timedelta(hours=1).total_seconds() returns 3600.0. In JavaScript you'd do 60 * 60 or 60 * 60 * 1000 for milliseconds. In Java, Duration.ofHours(1).getSeconds() returns 3600. These are all correct and all agree because the libraries define the hour as 3,600 seconds at the base. Where it gets messy is with units conversion libraries that infer hours from a timestamp string with a time-zone offset, because a timestamp with an offset sometimes gets interpreted as a wall-clock time rather than a duration. The library may subtract or add time-zone changes and give you 3,600 plus or minus an hour.
The practical rule is a split one: if you have a duration, multiply by 3,600; if you have a wall-clock time or a timestamp, convert to a duration before multiplying. If time zone changes matter in your data, use a library that distinguishes a Duration from a Timestamp (recent Java java.time and Spring's Duration do this; many simpler TimeSpan implementations don't). The arithmetic is rarely wrong. What's wrong is that two people, one working with a duration and one with a timestamp, both multiplied by 3,600 and got different answers because the inputs differed.
A Month Has No Fixed Second Count; You Derive It From 86,400, 604,800, and Days-in-Month
This is the edge case that causes real confusion in billing and scheduling: a month doesn't have a fixed number of seconds. A 30-day month is 2,592,000 seconds (30 × 86,400), a 31-day month is 2,678,400 seconds, and February in a common year is 2,419,200 seconds from 28 × 86,400. If a contract, bill, lease, or cloud data plan is tiered per month, "30 days" and "31 days" can differ by 86,400 seconds or more. That variation is why hourly-to-monthly conversions in pricing or compensation need to name the specific month or the specific number of days; any figure that assumes a single monthly second count is making a guess.
The leap-year differences add a fourth variant. February in a leap year has 29 days, which is 2,505,600 seconds. If you're building a billing system or converting automated scheduled durations, you have to decide up front whether your "month" is a calendar month, a fixed number of days, or a rolling 30-day period. These are three different durations and they'll produce different totals. The 3,600-second hour feeds into each one, but once the day count is set, the arithmetic changes.
The same principle applies at any larger scale. A year of 365 days is 31,536,000 seconds, and a leap year of 366 days is 31,622,400 seconds. Neither figure is a natural constant; they're the product of 3,600 × 24 × the actual day count. If you're doing time-zone-sensitive computations across a month boundary, you can also get another hour per DST transition on either side, which can change the second total. That's why so many systems default to fixed 30-day billing periods instead of calendar months.
For the question of how many seconds are in an hour, the answer remains 3,600 regardless of which of these larger units you're computing toward. The lesson is that downstream conversions inherit the variability of the larger unit. If you're computing a month in seconds, name the month. If you're computing a year in seconds, name the year type. If you're computing an hour, you already have the exact answer, and it doesn't need a day or a calendar to be true.
Short Durations Compare Cleanly Once You Divide by 3,600 and 86,400
Quick conversions are where a memorized 3,600 really earns its keep. If someone says a job takes 90 minutes, that's 5,400 seconds — a straightforward multiply. If a report says a session averaged two hours, that's 7,200 seconds. If a log shows 4,800 seconds, that's an hour and 20 minutes. The conversions run in both directions: divide by 3,600 to go from seconds to hours, multiply to go the other way. The table below gives you the common fractions and the arithmetic behind each, so you can verify the pattern and then reconstruct any other duration.
| Duration | Seconds | How to get there |
|---|---|---|
| 15 minutes | 900 | 15 × 60 |
| 30 minutes | 1,800 | 30 × 60 |
| 45 minutes | 2,700 | 45 × 60 |
| 1 hour | 3,600 | 60 × 60 |
| 1.5 hours | 5,400 | 1.5 × 3,600 |
| 2 hours | 7,200 | 2 × 3,600 |
| 4 hours | 14,400 | 4 × 3,600 |
The 15-minute figures are the ones that show up most in meeting and task planners. Three fifteen-minute chunks make one 2,700-second 45-minute slot, and the 3,600 hour is just the sum of fifteen-minute and 45-minute chunks laid together. That's not a coincidence; it's why the quarter-hour is such a common scheduling unit. The base-60 structure of the hour divides neatly into quarter-hour segments, which makes the arithmetic easy to do mentally and easy to verify on the back of a napkin.
For the conversions that don't come out to whole numbers, the formula is still the same. A 20-minute task is 1,200 seconds. A 7.5-minute task is 450 seconds. A 90-second task, going the other way, is 0.025 hours and 1.5 minutes. All of these drop out of the 3,600 anchor and the 60-second minute anchor. If you ever lose the formula, dividing by 3,600 gives you the decimal hour, and multiplying by 60 gives you minutes. That pair of moves covers every common direction of conversion.
The only real gotcha in short-duration conversions is mixing minutes and decimal hours. A time log that says "1.25 hours" means 1 hour and 15 minutes, which is 4,500 seconds, not 3,750. The decimal point sits in a different place depending on whether you're working in minutes or fractional hours, and getting the units straight is more important than the arithmetic. Once the unit is unambiguous, 3,600 does the rest.
A Standard Hour and the Solar Hour Differ by Seconds, Not by Thousands of Seconds
An hour on the clock is a fixed civil time unit, and it's exactly 3,600 seconds. An hour defined by the sun's apparent motion is a broader astronomical concept — the solar day is about 365.2422/366.2422 mean solar days, which is roughly 23 hours and 56 minutes and 4 seconds of sidereal time between successive transits of a fixed star. That's not a contradiction of the 3,600-hour; it's a different unit with a different physical meaning. If you're synchronizing a telescope or a satellite, you might use the sidereal hour (which is about 3,590 SI seconds, not 3,600). If you're scheduling a meeting, you use the civil hour and 3,600 is exact.
The most useful thing to know is that the confusion isn't a flaw in anyone's arithmetic; it's a domain difference. The civil hour is defined by the SI second, and 60 minutes times 60 seconds gives 3,600. Astronomers sometimes work in sidereal time because that's the frame in which the Earth's rotation against the stars is stable; the result is that a sidereal day is about 23 hours 56 minutes 4 seconds of solar time, and a sidereal hour is correspondingly shorter than 3,600 seconds. If you work a sidereal question with a 3,600-second hour, you'll be off by a handful of seconds per hour, which matters only in that specific domain. For nearly every reader, the civil hour is the one in use.
The same distinction explains why time-zone-aware computing is tricky in ways raw arithmetic isn't. A program that stores a timestamp in UTC and converts to local time using a fixed 3,600-second multiplier will be wrong on DST boundaries; a program that uses a proper time-zone library will be right. If you're mixing time zones and need a precise second count between two timestamps, use the library's elapsed-time function; let it handle the offset changes. The 3,600 figure applies to the definition of the hour, but elapsed civil hours across jumps are a different computation.
If you need a stable reference for the offset between civil and astronomical time, the IERS publishes data on the divergence and the bulletins that govern any leap seconds. That reference is the authority for the current state, and it's the place to look if a task actually depends on the distinction. For everything else, the 3,600-hour stands.
You Already Hold the Numbers That Make Hour, Day, and Year Conversions Reliable
The value of the 3,600 figure is leverage. One memorized number lets you rebuild 86,400 for a day, 604,800 for a week, and the second counts for shifts, billing periods, and logs without a reference table. That's the case for committing it to memory. It's exact, it's consistent across every computing platform and every timekeeping authority, and it's been stable for long enough that the 2035 leap-second phase-out plan is the only scheduled change. If you're going to hold a handful of time constants, this is the one that pays for itself fastest.
Mixed-unit conversions cause more real errors than slightly wrong arithmetic in single units. If a time log records "3,600" and you treat it as 3,600 minutes rather than 3,600 seconds, you're off by a factor of 60. The arithmetic was right; the unit was wrong. So the practical advice is to name the unit at every step and keep the conversion factor visible in the formula. If your spreadsheet shows 3,600 with no unit next to it, that's a warning sign as much as a value.
For anyone building or maintaining software that involves time, the right practice is to let a library handle the conversion and store durations in seconds at the boundary. That way the 3,600 factor is defined once, in one place, and everything else is a consistent multiplication from there. Once that base is stored safely, the downstream arithmetic is reliable, and most time bugs disappear before they start.
The exact number in one sentence: an hour is 3,600 seconds, and any conversion that respects the 60-second minute, the 60-minute hour, the 24-hour day, and the actual day count you're working with will land on the right figures.
FAQ
Does an hour ever have 3,599 or 3,601 seconds?
Yes, a UTC hour can contain 3,599 or 3,601 seconds when a leap second is inserted or removed, but such hours are extremely rare. A standard, non-leap hour is exactly 3,600 SI seconds. You'd only notice a leap-second hour in a precision-sensitive system, not in daily life. The IERS is the body that announces these adjustments.
Is it always 3,600 seconds whether I use Excel, Python, or a calculator?
For a duration of one hour, yes; all of them compute 60 × 60 and return 3600 (or 3600.0 in floating-point form). The exception is when you feed a wall-clock timestamp instead of a duration, because a spreadsheet or library may interpret the timestamp as a fraction of a day. Multiply a plain hour count by 3,600 and the answer will match across every platform.
How many seconds are in a day, a week, and a year?
A standard 24-hour day is 86,400 seconds, a 7-day week is 604,800 seconds, and a 365-day year is 31,536,000 seconds. A leap year of 366 days is 31,622,400 seconds. Build each one from the 3,600-second hour and the actual day count, and you'll never be off by more than the leap-second allowances.
Why is an hour 3,600 seconds and not 3,600.5?
The second is the base unit of time in the International System of Units, and the minute and hour are defined as exact multiples of it (60 and 3,600). There's no fractional remainder built into the definition. The wobble in the Earth's rotation gets handled by the leap-second system, not by changing the defined length of an hour.
How many seconds is a 30-minute meeting versus a 2-hour meeting?
Thirty minutes is 1,800 seconds (30 × 60), and two hours is 7,200 seconds (2 × 3,600). A 90-minute meeting is 5,400 seconds. Once you know the hour is 3,600 seconds, these comparisons are a simple multiplication and don't need a conversion tool.
Is 3,600 seconds per hour the same in every country?
Yes, the SI hour is universal; time zones change the offset from UTC, not the length of the hour. A DST "spring forward" day is a 23-hour day and a "fall back" day is a 25-hour day, but each individual hour in a non-leap system is still 3,600 seconds. If you're computing elapsed time across a time-zone jump, account for the offset separately.
How many seconds are in a 40-hour workweek?
A 40-hour workweek is 144,000 seconds, which comes from 40 × 3,600. That figure is useful for computing uptime budgets, payroll totals, or the number of seconds a system should be operational in a week. If your week has DST changes in it, the civil-time total may differ by 3,600 seconds depending on which way the shift goes.