How to Calculate Employee Attendance Percentage (With Formula and Examples)

Attendance percentage sounds like the simplest metric in HR: days present divided by days expected, times a hundred. In practice, most of the disagreements about an employee's attendance number don't come from the formula — they come from disagreeing about what counts as a 'day expected' in the first place. Get that part wrong and the percentage is confidently calculated and confidently wrong.

The basic formula

Attendance percentage = (days present ÷ total working days in the period) × 100.

For an employee present 19 days out of 22 scheduled working days in a month, that's (19 ÷ 22) × 100 = 86.4%. Straightforward, as long as both numbers in that fraction are counted consistently — which is where things usually go wrong.

What counts as a 'working day' — and what doesn't

Total working days should reflect the employee's actual assigned shift pattern, not a blanket 'weekdays in the month' assumption. An employee on a Sunday-to-Thursday schedule has a different set of working days than one on Monday-to-Friday, and calculating both against the same generic calendar produces a wrong number for at least one of them.

Company holidays should be excluded from the denominator entirely, not counted as an absence and not counted as a present day either — they simply aren't working days. The same logic applies to any day the business was fully closed for a reason unrelated to the individual employee.

How approved leave should be treated

This is where policies vary the most, and it needs to be decided explicitly rather than left to whoever's building the report that quarter. Common approaches: exclude approved leave days from the denominator entirely (so attendance percentage reflects only days the employee was expected to physically show up), or count approved leave as present (so the metric reflects overall reliability, treating a properly requested absence differently from an unexplained one).

Either approach is defensible. What isn't defensible is switching between them depending on which number makes a given month look better — the rule needs to be fixed and stated, ideally in the same attendance policy that defines grace periods and shift patterns.

Partial days and late arrivals

A day where an employee checked in three hours late is a different situation from a full absence, but a simple present/absent count treats them identically. Some companies track a separate 'punctuality' metric for late arrivals rather than folding it into attendance percentage, which keeps the headline number meaningful — 90% attendance with frequent lateness tells a very different story than 90% attendance with none, and collapsing them into one number hides that difference.

A worked example with leave excluded from the denominator

Take a 22-working-day month where an employee took 2 days of approved leave, was absent without approval for 1 day, and was present for the remaining 19. Excluding the leave days from the denominator: 19 present ÷ 20 expected working days (22 minus 2 leave days) = 95%. Including leave as present instead: 21 ÷ 22 = 95.5%. The two methods land close in this example, but they diverge more sharply the more leave an employee takes — which is exactly why the method needs to be picked in advance, not chosen after the fact.