Biweekly vs Semimonthly: How to Choose a Pay Schedule

I picked semimonthly for my first payroll because the dates were tidy. The 15th and the last day, two runs a month, everything closing on the month end alongside the books. It was the right answer for the team I had at the time, which was three salaried people who never worked an hour of overtime between them. It became the wrong answer about nine months later when I hired somebody hourly, and I did not notice for two more pay periods.

The thing nobody told me is that this is not a preference question. Biweekly and semimonthly sound like the same schedule described two ways, and almost every comparison you read treats them that way: one is 26 checks, the other is 24, pick whichever suits your cash flow. That framing hides the part that actually costs money. A biweekly pay period contains two whole workweeks. A semimonthly pay period contains a fragment of one, then some whole ones, then a fragment of another, and the fragments move every month.

So this is the comparison decided on mechanics rather than on taste: what 26 periods against 24 does to the gross on each check, exactly how the workweek arithmetic breaks under semimonthly and what you owe when it does, how benefit deductions land across both, what the third paycheck months and the 27 period year cost, what each schedule costs to process, and how to change schedules without anybody waiting longer for money they already earned. I build the people and records side at FirstHR, an onboarding and HR platform rather than a payroll provider, so none of this is a pitch. This is general information and not tax or legal advice.

What Actually Separates the Two Schedules

One is defined by an interval and the other by a calendar date, and every other difference follows from that. Biweekly repeats every fourteen days regardless of what month it is. Semimonthly repeats on two dates each month regardless of what weekday they land on.

Put plainly, biweekly is even in time and uneven on the calendar. Semimonthly is even on the calendar and uneven in time. Neither one is even in both, and choosing between them is really choosing which kind of unevenness you would rather absorb.

Two consequences of that do all the damage, and the rest of this article is really about them. Whole workweeks inside a period is the compliance consequence. Alignment with the month end is the accounting consequence, and it is why finance people reach for semimonthly while payroll people reach for biweekly. Both groups are right about their own problem.

26 Checks Against 24, and What Each One Is Worth

Dividing the same salary into 26 pieces instead of 24 makes each piece 7.69 percent smaller, or to put it the other way round, the semimonthly check is 8.33 percent larger than the biweekly one. The annual total does not move by a cent. Only the size and the timing of the pieces change.

Take a $60,000 salary. On semimonthly it is $60,000 divided by 24, which is $2,500.00 per check, twenty four times. On biweekly it is $60,000 divided by 26, which is $2,307.69 per check, twenty six times. The gap on each check is $192.31, and it is the single most misread number in this entire decision.

The hourly rows deserve a second look, because they are where semimonthly stops being tidy. A biweekly period at forty hours a week is 80 hours, in every period, in every month, forever. A semimonthly period runs ten to twelve working days depending on where the weekends fall, and as few as nine in the back half of a short February, so the same full time employee turns in 72, 80, 88 or 96 hours and the gross moves by $528.00 between the shortest period and the longest.

That variation is not a rounding problem, it is the correct answer. Hourly employees are paid for the hours they actually worked, and the number of available working days in a semimonthly period genuinely differs. What it removes is the ability to eyeball a payroll register and spot an error, because there is no constant to compare against. On biweekly, an 80 hour full time employee showing 72 hours is a question. On semimonthly, 72, 80, 88 and 96 are all plausible and you have to check the calendar to know which.

The salaried side is the mirror image. A salaried employee gets the same figure on every check under either schedule, so the divisor is the only difference, and the smaller biweekly check is easier to defend if you never let anybody discover it by surprise.

Why the Workweek Fits One Schedule and Cuts Across the Other

A biweekly pay period aligned to your workweek contains exactly two complete workweeks. A semimonthly period ends on a calendar date, and a calendar date lands on a different weekday every month, so the boundary falls in the middle of a workweek roughly twenty one times out of every twenty four.

The federal workweek is a fixed and regularly recurring period of 168 hours, seven consecutive twenty four hour periods. You choose when it starts, it does not have to be Sunday, and once you set it you leave it alone. Overtime for non-exempt employees is calculated inside that unit and nowhere else.

The regulation is blunt about the consequence. Under 29 CFR 778.104, the Act takes a single workweek as its standard and does not permit averaging of hours over two or more weeks, and the rule applies whether the employee is paid on a daily, weekly, biweekly, monthly or other basis. An employee who works 30 hours one week and 50 the next is owed overtime on ten hours, even though the two weeks average exactly forty.

Read that against the semimonthly picture and the problem appears immediately. Your pay period is not the unit that decides overtime. The workweek is. When the period boundary lands mid-week, the period contains part of a workweek whose total you cannot know until the following period opens, which means the period is not a self-contained payroll event any more.

Count how often this happens. There are 24 boundaries a year on a semimonthly schedule, at the 15th and the last day of each month. Each of those dates falls on an essentially arbitrary weekday. About three of the twenty four land on the last day of your workweek by coincidence. The other twenty one cut a week in half.

Overtime Across a Period Boundary, Worked Through

The method that works is to pay every hour at straight time in the period where it falls, then pay the overtime premium on the payday for the period in which that workweek ends. That is not a workaround, it is what the regulation directs, and it keeps the arithmetic auditable.

29 CFR 778.106 states that overtime compensation earned in a particular workweek must be paid on the regular pay day for the period in which such workweek ends, and that when the correct amount cannot be determined until some time after the regular pay period, the requirement is met if the employer pays the excess as soon after the regular pay period as is practicable. It adds that payment may not be delayed beyond the next payday after the computation can be made.

So take the split week from the diagram above. The workweek runs Sunday the 13th to Saturday the 19th, the semimonthly period closes on Tuesday the 15th, and an employee earning $22 an hour works ten hours a day on Sunday through Thursday. Fifty hours in the workweek: forty at straight time and ten at time and a half.

Check the total against the ordinary calculation and it agrees exactly. Forty hours at $22.00 is $880.00, ten hours at $33.00 is $330.00, and the sum is $1,210.00. The money is right. What has changed is that it arrived on two checks, the premium landed a period after three of the days that produced it, and somebody had to remember to go back and calculate it.

Run the same week on a biweekly schedule and there is nothing to remember. The entire workweek sits inside one period, the calculation happens once, $1,210.00 lands on one check, and the pay stub tells a story an employee can follow.

Two failures come out of this section more than any others. The first is computing overtime from period totals, which produces a number that is sometimes too high and sometimes too low and is never defensible. The second is treating the short piece of a split week as a quiet week and never going back, which underpays the premium and produces exactly the kind of finding a payroll audit is built to surface. Neither is possible on biweekly, which is the whole argument in one sentence.

Benefit Deduction Arithmetic Across 24 Periods and 26

Semimonthly wins this one outright, and the reason is that insurance carriers bill by calendar month. Twenty four deductions map onto twelve monthly invoices with no arithmetic at all. Twenty six do not map onto anything, and you have to pick which compromise you prefer.

Start with one employee paying $450.00 a month toward a family medical premium, $5,400.00 across the year. On semimonthly that is $225.00 a check, every check, and each calendar month collects exactly $450.00. The deduction register matches the invoice line for line, every month, without an adjusting entry.

Biweekly forces a choice. Spread the $5,400.00 across 26 checks and you get $207.69 each, which is even and easy to explain, but ten months of the year collect $415.38 against a $450.00 bill and two months collect $623.07. Twenty six checks at that figure come to $5,399.94, six cents short of the premium on the rounding, and no single month ever agrees with the invoice.

The alternative is to take $225.00 on the first two paydays of every month and take nothing on the third payday in the two months that have one. Now the invoice reconciles monthly, but the employee gets two checks a year with a noticeably bigger net and your payroll setup has to identify the skip periods correctly for the life of the schedule.

Then there are the deductions that are annual caps rather than monthly bills, and those behave differently again. A health flexible spending account election of $2,600 against the 2026 limit of $3,400 (Revenue Procedure 2025-32) comes to $108.33 a period across 24 and exactly $100.00 across 26. Both are fine, provided the divisor is the true number of remaining periods and not the nominal one, because the election has to be fully collected by the end of the plan year.

Retirement is where this bites hardest. A percentage election self-corrects: three percent of gross is three percent whichever way you slice the year. A flat dollar election does not. The 2026 elective deferral limit is $24,500 (IRS Notice 2025-67). An employee electing a flat $1,020 a period reaches $24,480 across 24 semimonthly checks and stops just under the limit. The same $1,020 across 26 biweekly checks reaches $26,520, which is $2,020 over, and somebody has to catch it before the year closes.

Garnishments carry a version of the same issue. The federal exemption floor is expressed per pay period and the multiplier changes with pay frequency, so moving from 24 periods to 26 changes the protected amount on every order you are administering. It is a per-order recalculation rather than a global setting.

The Third Paycheck Months and the 27 Period Year

Biweekly carries two calendar surprises that semimonthly does not have at all. Twenty six paydays spread across twelve months means ten months with two paydays and two months with three. And because a year is not an exact multiple of fourteen days, a 27th payday eventually lands inside one calendar year.

The three payday months are the ordinary one and they arrive every year. Which two months they are depends on where your first payday of the year sits, and it moves. If you budget payroll as two runs a month, twice a year the cash requirement rises by half again, and it lands in months you did not choose. On a $60,000 salary that is an extra $2,307.69 per person in each of those two months.

People sometimes call those extra checks free money and they are not. Nobody is paid more across the year. What happens is that the cash leaves the business in a lumpier pattern, and the businesses that get caught are the ones running a tight operating account against a payroll line built on twenty four runs.

The 27 period year is the rarer and larger one. A year holds 365 days, which is 26.07 fourteen day periods, so a biweekly cycle drifts by about a day a year and two in a leap year until a 27th payday falls inside a single calendar year. Divide salaries by 26 and run all 27 and you have paid 27 twenty sixths of every salary, which is 3.85 percent more payroll than you budgeted.

Semimonthly has neither problem. It is 24 periods in every year, leap or not, two per month without exception, and the monthly payroll figure you budget in January is the monthly payroll figure in every other month. That predictability is the single structural advantage semimonthly holds that nothing can take away from it.

Processing Cost and the Administrative Load

Biweekly costs you two extra payroll runs a year, which is roughly eight percent more processing events, and if you pay per run that shows up as a line item. Semimonthly costs you overtime complexity and a moving payday, and those show up as time and as risk rather than as an invoice.

Count what a run actually costs before assuming the fee is the whole story. Every run means timesheet approval, a register review, funding the account, submitting the file and answering the questions that follow. Two extra cycles a year is two extra rounds of all of it. For a small employer that is a real number of hours, and it recurs forever.

Now count what semimonthly costs on the other side. The payday moves across the week, so the banking cutoff for direct deposit moves with it. A payday that lands on a Monday needs the file submitted before a weekend, and that submission deadline changes shape twenty four times a year rather than never. This is where late funding incidents come from.

There is also the month end question, which almost nobody raises and which finance teams care about a great deal. A semimonthly period running from the 16th to the last day of the month closes exactly on the month end, so the month sits inside whole pay periods and nothing has to be prorated across the boundary. A biweekly period almost never closes on the last day of the month, so at every month end you are accruing the wages earned between the last period close and the month end, twelve times a year, forever.

That accrual is not hard, it is just permanent. It is a recurring journal entry that has to be right, reversed and re-run every month, and it is the reason a bookkeeper will push for semimonthly while your payroll administrator pushes for biweekly.

Employee Cash Flow and What People Actually Notice

Semimonthly gives employees the same amount every calendar month, which matches how rent, mortgages and loan payments work. Biweekly gives a smaller amount at a perfectly even fourteen day rhythm, plus two months a year that feel like a windfall. Neither is better in the abstract and both are noticed immediately.

The point people miss is that semimonthly is even on the calendar and uneven in time. Paydays on the 15th and the last day of the month sit sixteen days apart from the 15th to the 31st, fifteen days apart from the 31st to the 15th, and only thirteen days apart across February. The interval swings between thirteen and sixteen days even though the dates never move.

Biweekly is the opposite and it is genuinely even: fourteen days, always, the same weekday, forever. For anyone budgeting week to week that regularity is worth more than the tidiness of a fixed date, which is why hourly workforces tend to expect it and why some industries have never used anything else.

Then there is the weekend and holiday rule, which is not optional and which only affects one of the two schedules meaningfully. A biweekly payday on a Friday only moves for a bank holiday. A semimonthly payday on the 15th lands on a weekend several times a year and has to be moved, which means your written policy needs to say in advance whether it moves earlier or later, and moving it later can collide with a state deadline.

The last piece is perception. Moving from semimonthly to biweekly cuts the visible number on the check by 7.69 percent, and a meaningful share of people will read that as a pay cut no matter what the annual figure says. That is a communication problem rather than a payroll problem, but it is the one that generates the meetings, so plan the announcement with the same care as the calculation and make sure the pay stub shows enough detail to answer the obvious question.

Hourly Staff Decide This More Often Than Salaried Staff

If a meaningful share of your payroll is non-exempt and hourly, choose biweekly and stop deliberating. If everybody is exempt and salaried, both schedules work and semimonthly has the better argument. The composition of the workforce settles this faster than any other input.

The hourly case for biweekly has three independent legs and each one stands on its own. Overtime lands inside the period, so no calculation crosses a boundary. The period is always exactly 80 hours at full time, which gives you a constant to check the register against. And the timesheet approval window is the same length every period, so the operational rhythm never changes.

The salaried case for semimonthly is quieter but real. There is no overtime to align, the check is the same figure either way, the periods close on the month end so the accounting is clean, and benefit deductions map onto carrier invoices without a workaround. If you have no non-exempt staff, the entire compliance argument against semimonthly evaporates.

One shortcut deserves a specific warning. Some employers on semimonthly assign hourly staff a fixed 86.67 hours a period, which is 2,080 hours divided by 24, so the check looks constant. That pays an average instead of actual hours and understates pay in the long periods. Pay real hours from real timesheets, every period.

Exempt staff are unaffected by the choice in a legal sense. The salary basis test requires a predetermined amount paid on a weekly or less frequent basis and not subject to reduction for variations in the quality or quantity of work, and both schedules satisfy that comfortably.

Mixed workforces raise the obvious question, and the answer is that you may run both. Hourly staff biweekly and salaried staff semimonthly is legal, it is not unusual, and it solves the real problem. It also gives you 50 payroll events a year, two calendars, two deduction divisors, two funding rhythms and two of every reconciliation, which for an employer without a dedicated payroll person is usually a worse trade than picking one schedule and living with it.

The State Laws That Narrow the Choice Before You Make It

Check this first, because it is the only input that can remove an option outright. Federal law sets no pay frequency requirement at all. Frequency is state law, and several states rule one of these two schedules out for at least part of the workforce.

The Department of Labor maintains a state payday requirements table, last revised in January 2023, which is the right starting point and not the finishing one. The statutes move, and the version that governs you is the current text from the state where your employee physically works, which is a different question from where your business is registered.

Two patterns are worth extracting from that table. Some states restrict frequency by worker type rather than across the board, so the same employer can be free to choose for one group and constrained for another. And some states pair a frequency rule with a lag rule, which is the deadline for paying wages after the period ends, and the lag rule can bind harder than the frequency one.

Multi-state employers have to satisfy the strictest rule that reaches any single employee, and a remote hire can change the answer for everybody. One person in a state requiring weekly pay for their occupation can take your preferred schedule off the table for that person permanently.

Nothing here is legal advice and none of it substitutes for the current text of your own state statute. Read it before you announce a schedule rather than after, and re-read it whenever you hire into a new state, because payroll compliance failures in this area are cheap to prevent and expensive to unwind.

Switching Schedules Without Shorting Anyone

You can change pay schedules prospectively as long as nobody waits longer for wages already earned than their state permits, you give whatever advance written notice the state requires, and no earned wages are reduced or delayed to make the calendar work. The failure mode is always the same, and it is always the transition rather than the destination.

What goes wrong is that an employer moving between schedules quietly stretches the changeover so people wait an extra week or two for money they have already worked for. That is not a scheduling adjustment, it is a delayed wage payment, and several states measure it exactly that way. The correct treatment is a bridge period paid on its own terms.

Direction matters for the announcement more than for the mechanics. Moving from semimonthly to biweekly drops the visible check by 7.69 percent and needs the most careful communication of anything in this article. Moving from biweekly to semimonthly raises it by 8.33 percent and sells itself, but it is the direction that introduces the overtime complexity, so the internal work is heavier even though the conversation is easier.

One detail people forget in either direction: whatever is in flight has to keep working. Year to date totals carry across, the Social Security wage base of $184,500 for 2026 is tracked on the year rather than the period, retirement contributions accumulate against an annual limit, and a mid-year switch does not reset any of it. Reconcile year to date figures for every person on the first run under the new schedule, and treat anything that does not tie out as a stop rather than a rounding issue.

Frequently Asked Questions