End of Service in Saudi Arabia: Formula, Resignation Rules and Worked Examples

Under Article 84 of the Saudi Labour Law, the end-of-service award is half a month's wage for each of the first five years and a full month's wage for each year after, calculated on the last actual wage (allowances included unless the contract agrees otherwise), with fractions of a year pro-rated. Article 85 cuts the award on resignation: no award under two years, one-third for 2-5 years, two-thirds for 5-10 years, full at ten years. Settlement is due within one week (two weeks if the worker ended the contract). That paragraph is the entire economics of Saudi Arabia end-of-service: read it twice before modelling anything. Unlike the UAE's single 21/30-day rule, several Gulf states scale the award by resignation behaviour (Saudi Arabia's resignation treatment is above) and compute on different wage bases - Saudi Arabia uses the base described above, not the UAE's basic-wage concept. The table below puts the formula in one view.

Apply the formula above to a concrete case: take the employee's service years, apply the per-year rate for each tranche, pro-rate any partial year, and check any cap. Then re-run it with the resignation variant if the exit is employee-initiated - in Saudi Arabia, unlike the UAE, the exit reason can materially change the number. Keep the worksheet: end-of-service disputes in Saudi Arabia are arithmetic disputes, and the employer with the worksheet wins them. Confirm the worked figures against the current statutory text before using them in a settlement.

Under Article 84 of the Saudi Labour Law, the end-of-service award is half a month's wage for each of the first five years and a full month's wage for each year after, calculated on the last actual wage (allowances included unless the contract agrees otherwise), with fractions of a year pro-rated. Article 85 cuts the award on resignation: no award under two years, one-third for 2-5 years, two-thirds for 5-10 years, full at ten years. Settlement is due within one week (two weeks if the worker ended the contract). The wage base is where Saudi Arabia diverges most from UAE practice: several Gulf states compute on a broader 'actual wage' including allowances, where the UAE computes strictly on basic wage. Read the contract's wage definition against the statute before calculating - and when structuring new contracts in Saudi Arabia, understand that every allowance folded into the wage base compounds into the end-of-service liability for the entire tenure.

End-of-service dues must be settled promptly at the end of the employment relationship - confirm Saudi Arabia's exact statutory deadline. Pay through documented banking channels with a signed settlement statement itemising each component. If the employer defaults, the route runs through the Ministry of Human Resources and Social Development (MHRSD)'s complaint mechanism and then the competent court or tribunal. Employees should keep the contract, payslips and the settlement worksheet; employers should keep the same plus proof of payment.

Disputes go through the Ministry of Human Resources and Social Development (MHRSD)'s settlement process first, then adjudication. The employer's defences are narrow: prove the service period, prove the wage base, prove payment. The employee's claim is arithmetic. This asymmetry is why the worksheet-and-receipt discipline matters more than legal argument. For groups, align the Saudi Arabia dispute playbook with the UAE one - early settlement of clear-cut arithmetic claims is cheaper than defending them in both countries.

UAE: 21 days of basic wage per year for five years, 30 days after, two-year cap, reason-independent since 2022. Saudi Arabia: Under Article 84 of the Saudi Labour Law, the end-of-service award is half a month's wage for each of the first five years and a full month's wage for each year after, calculated on the last actual wage (allowances included unless the contract agrees otherwise), with fractions of a year pro-rated. Article 85 cuts the award on resignation: no award under two years, one-third for 2-5 years, two-thirds for 5-10 years, full at ten years. Settlement is due within one week (two weeks if the worker ended the contract). The planning consequences: Saudi Arabia provisions must model the resignation haircut where it exists, use the local wage base, and track any cap or transitional rules (Bahrain's March 2024 SIO shift and Oman's July 2023 cutover being the live examples). Run two provision models; a blended 'GCC gratuity rate' is wrong in both countries.

Accrue the Saudi Arabia end-of-service liability monthly per employee, recompute at each month-end, and reconcile the balance-sheet provision to the individual schedules. For Saudi Arabia this means modelling the resignation-adjusted expected value where haircuts apply - or provisioning the full formula conservatively and releasing on resignation. Either way, the provision is a real liability that auditors and buyers diligence: an unprovided Saudi Arabia workforce is a purchase-price adjustment waiting to happen.

An annual self-audit of Saudi Arabia payroll costs little and catches what monthly routine misses: the findings that compound silently between inspections. Run five tests. First, wage-protection filing completeness: every pay cycle filed, every rejection resolved, the register matching headcount - pull the submission history and match it to the journals, no sampling, the data is digital. Second, social-insurance computation: recompute employer and employee shares for a stratified sample on the statutory base, and confirm the national/expatriate split was applied correctly - misclassified employees are the classic expensive finding. Third, end-of-service provision: recompute per employee on the local formula and reconcile the balance-sheet provision to individual schedules. Fourth, contract-to-payroll match: the registered wage vs the payroll wage for a sample - drift here corrupts every downstream computation. Fifth, records: the monthly evidence pack exists and is retrievable. Document the method as you go so a stranger could re-perform every test; an audit nobody could repeat is an opinion, not an audit. Where a local provider runs the payroll, test their outputs against your source data - the contract says they are responsible, but the liability sits with you. Close the loop on every finding: each gets an owner, a remediation deadline, and a re-test in the next cycle proving the fix held. Findings without owners become next year's findings. Keep the audit file with the year's compliance records - it is the document that proves diligence if the regulator ever asks what you did to get payroll right. A clean self-audit is also the fastest possible answer to a due-diligence questionnaire in a funding round or acquisition: the five tests, the method, the findings, the closures, all on paper. Start this quarter even if the first run is rough - a rough audit beats a polished assumption, and the second year's run will be unrecognisable in its rigour.

Every Saudi Arabia payroll cycle should produce a retrievable evidence pack: the payroll journal, the wage-protection submission confirmation (or the local equivalent filing receipt), itemised payslips, input approvals, and the correction audit trail where adjustments were made. Alongside the monthly pack, keep the standing file per employee: the registered contract, identification and permit copies, wage-structure documentation, and the end-of-service computation history. Retention periods are set by local regulation - confirm the current Saudi Arabia requirement, and retain at least as long as the longest limitation period for labour claims. Store for retrieval, not for storage: organised by period and employee, indexed, with access controls on salary data. The test of a records system is a surprise query - a regulator's request, a dispute, a due-diligence data room - answered in hours from the file, not reconstructed over weeks from memory and bank statements. Employers who can produce the pack win disputes; employers who cannot, fund them. Build the pack into the monthly routine rather than reconstructing it: the payroll calendar should name the pack as a deliverable with the same status as the salary run itself. When the finance team asks why payroll administration costs what it does, the evidence pack is the answer - it is the difference between a salary spreadsheet and a defensible payroll operation. In a multi-entity group, standardise the pack's contents across entities so internal audit and external counsel see one format everywhere; consistency is what makes a regional payroll function credible rather than merely busy. Treat the pack as the product of payroll administration, not its by-product - the salary run is the event, the pack is the proof, and proof is what survives scrutiny.

Groups entering Saudi Arabia choose between three operating models, and payroll complexity differs sharply between them. The registered branch or subsidiary runs full local payroll: local contracts, local wage-protection filing, local social insurance, local end-of-service - the complete discipline this guide describes. Secondments (home-country employment, host-country work) look lighter but create the hardest payroll questions: which country's social insurance applies, how the wage-protection filing treats a seconded employee, where the end-of-service liability sits, and how the recharge is documented for tax. Get written advice on the secondment structure before the first payroll run - unwinding a wrongly structured secondment costs multiples of structuring it right. The employer-of-record (EOR) model outsources the local employment to a licensed provider: fastest to market, but the EOR's payroll quality becomes your compliance exposure, and exit means migrating employees to your own entity later. None of the three models avoids Saudi Arabia payroll rules; they only change who operates them. Whichever model you choose, keep a one-page model register per employee: employment vehicle, payroll operator, social-insurance jurisdiction, end-of-service owner, and wage-protection filing responsibility. Review the register whenever headcount changes or regulations move - operating models drift, and drift is how a secondee ends up on nobody's filing. The register takes an hour to build and prevents the multi-jurisdiction tangles that cost months to unwind. Revisit the model choice itself annually: the EOR bridge that was right for five explorers is wrong for fifty operators, and the secondment that made sense for a six-month project calcifies into a permanent compliance anomaly if nobody re-examines it.

Payroll is where small UAE employers bleed cash quietly: a mis-split basic wage compounds into a gratuity surprise years later, one missed WPS cycle triggers per-employee fines, and year-end brings accruals nobody booked. TaxAdvisors runs managed payroll for UAE employers - WPS salary files every cycle, gratuity provisions calculated on the real basic-wage split, MOHRE filing calendars, and bookkeeping that keeps payroll journals reconciled to the bank. If you would rather have this handled than studied, talk to us before the next pay run, not after the first fine.