We use cookies to understand which pages and tools are useful, nothing more — no ad tracking across other sites. You can change this later from the footer.

Skip to content
Lekhya.in

Checklist7 min readPublished

Payroll audit procedure: a complete checklist

The full sequence for a payroll audit: master data testing, gross-to-net recomputation on a sample, statutory deduction testing, and the approval trail a payroll run needs before it is paid.

A payroll audit procedure runs in four stages: test the employee master for unauthorized or stale records, recompute gross-to-net for a sample of employees across a sample of months, test statutory deductions (PF, ESI, professional tax, TDS) against the applicable rules, and verify that every payroll run carries evidence it was reviewed and approved before disbursement. Each stage depends on the one before it — a master data defect invalidates everything computed on top of it.

Why payroll audit is treated as a separate section

Payroll is usually one of the largest expense lines in a services or manufacturing business, and it is also one of the few expense lines where a single data error — a wrong rate, a duplicate employee code, an unauthorized master change — repeats itself every month until someone catches it. Testing a transaction sample without first testing the master that generates those transactions misses exactly this kind of error.

Stage 1: employee master data testing

The employee master is the table that drives every payroll run: employee code, name, bank account, PF/ESI applicability, salary structure, date of joining, date of exit. Changes to it should be rare, approved, and visible.

  1. 01Obtain the full employee master and the change log for the periodMost payroll systems log master changes with a timestamp and user ID. If the client's system cannot produce this log, note it as a control gap rather than skipping the step.
  2. 02Reconcile headcount to HR records independently of payrollCompare the payroll master count to the HR attendance system or the organization chart maintained outside payroll. A name on the payroll master that does not appear anywhere else is the definition of a ghost employee risk.
  3. 03Test for duplicate bank accounts and duplicate PAN/Aadhaar across employee recordsOne bank account number linked to two employee codes is the single highest-yield check in this stage. It surfaces both ghost employees and genuine but undisclosed dual employment.
  4. 04Trace a sample of new joiners to appointment letters and a sample of exits to the full-and-final settlementConfirms the master reflects real employment events and that exits actually stopped future payroll runs rather than being flagged and forgotten.
  5. 05Review master changes made just before a payroll run for approval evidenceA salary revision, bank account change, or new-joiner entry added in the window between attendance lock and payroll processing, with no corresponding approval, is exactly the pattern used to divert a payment.

Ghost employee risk is not limited to fictitious names. An employee who genuinely resigned but whose master record was never deactivated is operationally the same risk, and it is far more common. Exit processing discipline is worth testing on its own.

Stage 2: gross-to-net recomputation

This is the arithmetic core of the audit: for a sample of employees in a sample of months, recompute gross pay from the salary structure and attendance data, then recompute every deduction down to net pay, and compare to what was actually paid.

What a gross-to-net recomputation needs, line by line
ComponentRecompute fromCommon break
Basic, HRA, and other fixed componentsThe approved salary structure for that employee as of that monthA salary revision applied to the master but not reflected in the run being tested, or vice versa
Variable pay tied to attendance (LOP, overtime)The attendance or biometric system, independently of the payroll system's own attendance feedAttendance corrections applied in payroll that were never approved in the attendance system
PF, ESI, professional taxThe applicable statutory rate and wage ceiling for that employee and monthA stale rate table not updated after a notified change, or ESI continuing to be deducted for an employee whose gross crossed the applicability ceiling mid-year
TDS on salaryThe declared tax regime and investment declarations on file for that employeeTDS computed on an old declaration after the employee submitted a revised one
Net payGross less all deductions above, recomputed independentlyA manual adjustment applied directly to net pay with no trace back to any component above it
What a gross-to-net recomputation needs, line by line

Pick the sample across both employees and months, not just employees. A control that holds for March can fail in a month with a mid-cycle salary revision or an attendance correction, and a monthly sample that always lands on the same quiet month will never find that.

Stage 3: statutory deduction testing overview

Gross-to-net recomputation checks that deductions were calculated correctly against the employee master. Statutory deduction testing checks the other direction: that what was deducted and the matching employer contribution were actually deposited with the right authority, on time, for the right amount. This overlaps with, but is distinct from, PF/ESI compliance testing and TDS testing, which are detailed enough to warrant their own procedures.

  • Agree total PF and ESI deducted per the payroll register to the challans filed for the corresponding month.
  • Agree total TDS deducted to the amount deposited and, at a summary level, to the quarterly TDS return.
  • Test professional tax deduction against the applicable state slab for a sample of employees — slabs vary by state and change periodically.
  • Check due-date compliance for each statutory payment and flag any late deposit for the interest-on-delay implication.

Stage 4: the approval trail for payroll runs

A payroll run that is arithmetically correct but never reviewed by anyone before disbursement is still a control failure. The approval trail should show, for each run tested, who prepared it, who reviewed it, and what the reviewer actually looked at.

  1. 01Identify who has system access to initiate a payroll run and who has access to approve disbursementThe same person should not hold both unless the entity is small enough that segregation is genuinely impractical, in which case a compensating review is needed and should be evidenced.
  2. 02Obtain the sign-off evidence for the runs in your sampleAn email approving a lump-sum total is weaker evidence than a reviewer sign-off on the actual register, because the former cannot show that anything was actually checked.
  3. 03Trace the approved register to the bank file or salary disbursement instructionConfirms the amount approved is the amount actually paid, and that no change was made to the file between approval and transmission to the bank.
  4. 04Check that month-on-month variance in the total payroll cost was reviewed by someone, and that large variances have a stated reasonThis is the cheapest control an entity can run and the one most often skipped in practice.

Common practice, not a requirement

Running the headcount-to-cost variance analysis before sampling, rather than after, is not a requirement but it reliably improves where the sample lands. Months or departments with unexplained movement deserve a heavier weight in the sample than a purely random draw would give them.

How we think about this

Everything in stages one through three is a recomputation against a rule — a rate, a ceiling, a formula — and a computer applies that rule to every employee in every month with no fatigue and no sampling risk. What a reviewer in stage four decided when they signed off a payroll run is not something software can reconstruct after the fact; that judgement either exists in the file or it does not. The procedure above is deliberately ordered so the computable work is done first and at full population, leaving the judgement-based review where it belongs — with a person.

This is practical guidance written for audit teams, not an authoritative statement of any standard or statute. Standards, forms and timelines change. Check the current text issued by the ICAI and the relevant authority before relying on anything here in an audit file.

Free tools, no signup

Each one reads a file entirely in your browser — or, for two of them, just answers a few questions — and nothing is ever uploaded.