FMLA and Payroll: Calculating Pay During Intermittent Leave
7/28/2026
Intermittent FMLA leave is the hardest leave type to administer in payroll, and the difficulty is almost entirely in tracking. A continuous 12-week absence is a single event. The same entitlement taken as two hours here and half a day there, across a year, is hundreds of events that must each be recorded in the correct increment against a correctly calculated entitlement.
Get the increment wrong and you either exhaust an employee's entitlement early — which is interference — or fail to charge time that should have counted, leaving the employer with more leave to provide than the law requires.
The Entitlement Is Measured in Hours
The entitlement is commonly described as 12 weeks. For intermittent leave it functions as hours: 12 weeks multiplied by the employee's normal weekly schedule.
- A 40-hour employee has 480 hours
- A 30-hour employee has 360 hours
- A 20-hour employee has 240 hours
Two consequences that are frequently missed:
Overtime does not increase the entitlement, but mandatory overtime hours the employee would have been required to work and could not because of the leave generally do count against it. Voluntary overtime does not.
A variable schedule requires an average, generally computed over the 12 months preceding the leave. This must be recalculated for each new leave year rather than carried forward.
Our Leave Management Compliance Suite and Certified Leave Administrator cover the framework.
Track in the Smallest Increment You Use
The rule that generates most interference claims.
An employer must account for intermittent leave in increments no greater than the smallest increment it uses for other types of leave. If your timekeeping system tracks other absences in six-minute or fifteen-minute increments, you must track FMLA the same way.
You may not round FMLA usage up to a half day or a full day where you do not do so for other leave. Charging a full day for a two-hour medical appointment overstates usage, exhausts the entitlement prematurely, and is interference — a claim that is straightforward to prove from your own records.
A narrow physical-impossibility exception exists where an employee cannot begin or end work mid-shift for genuine operational reasons, but it is applied narrowly and should not be assumed.
The practical consequence: configure the FMLA increment to match your smallest leave increment, and verify it rather than assuming the system default is correct.
Choose and Document the 12-Month Method
Four permitted methods for measuring the 12-month period, and the choice materially affects availability:
Calendar year. Simple, and it permits an employee to take 12 weeks in late December and another 12 in January.
Any fixed 12-month period — a fiscal year, or an anniversary date. Same stacking issue at the boundary.
Measured forward from the first date leave is taken. The employee's year begins with their first use.
Rolling backward from each date leave is used. Prevents stacking, and is generally the most protective of the employer.
Two requirements: the method must be applied consistently to all employees, and changing methods requires advance notice with transitional protection for employees. Where an employer has not chosen a method, the one most beneficial to the employee generally applies — which means failing to decide is itself a decision, and an expensive one.
The Exempt Employee Trap
The highest-consequence payroll issue in intermittent leave.
Ordinarily, deducting from an exempt employee's salary for a partial-day absence violates the salary basis requirement and can destroy the exemption — potentially for everyone in the same job classification subject to the practice.
Unpaid FMLA leave is a specific exception. An employer may deduct for partial-day absences taken as unpaid FMLA leave without jeopardizing the exemption. This is one of the few permitted partial-day deductions.
Three cautions:
The exception is limited to FMLA-qualifying leave. A partial-day absence for a reason that does not qualify, or after the entitlement is exhausted, does not benefit from it.
The leave must be properly designated. An employer that never designated the absence as FMLA cannot rely on the exception for it.
Substituted paid leave is not an unpaid deduction. Where the employee uses accrued paid leave, they are paid in full and the salary basis question does not arise.
See our exempt vs. non-exempt guide.
Paid Leave Substitution
FMLA leave is unpaid. An employer may require, or an employee may elect, that accrued paid leave run concurrently — "substitution."
The rules that matter:
The employer may require substitution, but only subject to the terms of its own paid leave policy. If your PTO policy requires two weeks' notice, you generally cannot require substitution for an unforeseeable absence that could not satisfy it.
Where the leave is already paid — through workers' compensation, a disability plan, or a state paid family leave program — the ordinary substitution rules generally do not apply, because the leave is not unpaid. The employee's entitlement to supplement with accrued leave in that situation depends on the arrangement.
The employee must be notified of the substitution requirement.
Substitution does not extend the entitlement. Twelve weeks is twelve weeks whether paid or unpaid.
Where paid leave is exhausted, the remaining FMLA leave is unpaid but still job-protected.
For intermittent leave specifically, substitution means an employee taking two hours of FMLA may have two hours of PTO applied — which requires the payroll system to charge both the FMLA entitlement and the PTO balance for the same absence, in matching increments.
Benefit Deductions During Intermittent Leave
Group health coverage must continue on the same terms throughout FMLA leave, including intermittent leave.
For intermittent leave this is usually manageable, because the employee continues receiving some wages in most periods and deductions can be taken normally. The problem arises in a period where the employee's hours drop enough that wages do not cover the deductions.
Handle it the same way as continuous leave: pre-payment, pay-as-you-go, or catch-up recovery under a written agreement made in advance. Recovering advanced premiums from later wages is a deduction requiring specific authorization in most states, and arranging it retroactively is a weaker position. See our cafeteria plans guide and voluntary deduction authorization guide.
The Recurring Errors
- Rounding usage up to half or full days, exhausting the entitlement early
- Never designating the leave, so the entitlement did not run and the employee has 12 weeks still available
- No 12-month method chosen, defaulting to the one most beneficial to the employee
- Docking exempt salary for a partial-day absence not properly designated as FMLA
- Counting voluntary overtime against the entitlement
- Failing to recalculate a variable schedule for each leave year
- Attendance points assigned for FMLA-protected absences — an automated system that does not know which absences were protected
- Benefit premiums uncollected during a low-wage period, then recovered without authorization
- Failing to coordinate with state paid family leave providing concurrent wage replacement
- Requiring substitution where the leave is already paid
The attendance point issue deserves emphasis. If your attendance system assigns points automatically, protected absences must be excluded — and that is a systems configuration problem rather than a policy one. It is also among the most commonly litigated FMLA interference claims.
Payroll's Checklist
- [ ] Compute the entitlement in hours, from the employee's normal weekly schedule
- [ ] Recalculate the average for variable schedules each leave year
- [ ] Configure the FMLA increment to match your smallest leave increment
- [ ] Choose, document, and consistently apply a 12-month method
- [ ] Designate every qualifying absence as FMLA and notify the employee
- [ ] Charge only mandatory overtime hours missed against the entitlement
- [ ] Apply substitution per policy, charging both entitlement and PTO in matching increments
- [ ] Use the partial-day exception for exempt employees only where leave is properly designated and unpaid
- [ ] Arrange benefit premium collection in advance for low-wage periods
- [ ] Exclude protected absences from attendance point systems
- [ ] Coordinate with state paid family leave and disability wage replacement
- [ ] Retain FMLA records for the required period, separately from medical information
Auditing Your Intermittent Leave Configuration
Because the errors in this area are configuration errors rather than judgment errors, they can be found deliberately. Four checks.
Check the increment. Find your smallest leave increment for any other leave type, then confirm the FMLA increment matches it. Do not accept the system default — verify by taking a test absence of the smallest possible duration and confirming what was charged. This single check finds the most common and most costly error.
Check the entitlement calculation. For several employees, confirm the hours entitlement equals 12 weeks times their actual normal weekly schedule, and confirm that any variable-schedule average was recomputed for the current leave year rather than carried forward.
Check the 12-month method. Confirm one has been chosen, is documented, and is applied consistently. Where you cannot find a documented choice, assume none was made and take advice — the method most beneficial to the employee generally applies, which may mean employees have more entitlement available than your system shows.
Check the attendance system. Take a sample of employees with recorded FMLA usage and confirm no attendance points were assigned for those dates. This is the check most likely to find an active liability, because the exclusion is usually a manual step someone has to remember.
Run all four annually and after any system upgrade, since an upgrade can silently change increment handling or rounding behavior. Document what you tested and the result — it is evidence of good faith if an interference claim ever arises.
Frequently Asked Questions
How is FMLA intermittent leave tracked?
In hours rather than weeks — the entitlement equals 12 weeks multiplied by the employee's normal weekly schedule, so a 40-hour employee has 480 hours and a 30-hour employee has 360. Usage must be recorded in increments no greater than the smallest increment used for other leave types, and a variable schedule requires an average computed over the preceding 12 months and recalculated each leave year.
Can an employer round FMLA leave to the nearest half day?
Not where it does not round other leave types that way. Charging a full or half day for a two-hour absence overstates usage, exhausts the entitlement prematurely, and constitutes interference — a claim easily proven from the employer's own records. A narrow physical-impossibility exception exists where an employee genuinely cannot begin or end mid-shift, but it is applied narrowly.
Can an employer dock an exempt employee's pay for intermittent FMLA leave?
Yes — this is one of the few permitted partial-day deductions from an exempt employee's salary. Ordinarily such a deduction violates the salary basis requirement and can destroy the exemption for an entire classification, but unpaid FMLA-qualifying leave is a specific exception. It applies only where the leave actually qualifies and was properly designated, and it does not apply after the entitlement is exhausted.
Can an employer require employees to use PTO during FMLA leave?
Generally yes, subject to the terms of the employer's own paid leave policy — so a PTO policy requiring advance notice may not support requiring substitution for an unforeseeable absence. Where the leave is already paid through workers' compensation, a disability plan, or a state paid family leave program, the ordinary substitution rules generally do not apply because the leave is not unpaid. Substitution never extends the 12-week entitlement.
What 12-month method should an employer use for FMLA?
The four permitted options are the calendar year, any fixed 12-month period, measured forward from first use, or rolling backward from each date used. The rolling-backward method prevents an employee stacking 12 weeks at the end of one year and 12 at the start of the next, making it generally most protective of the employer. The method must be applied consistently — and where none has been chosen, the method most beneficial to the employee generally applies.
Can FMLA absences count against an attendance policy?
No, and this is among the most commonly litigated interference claims. Assigning attendance points for FMLA-protected absences is unlawful, and it is usually inadvertent because automated attendance systems do not know which absences were protected. Excluding protected absences is a systems configuration problem rather than a policy statement, and someone must own it.
Going Deeper
Verify that your system's FMLA increment matches your smallest leave increment rather than assuming the default is correct, document your 12-month method, and ensure protected absences are excluded from any automated attendance calculation.
