Guide · updated 16 September 2026

Why your payslip can differ from an annual tax estimate

PAYG withholding is collected through payroll. Your annual tax assessment uses the income and circumstances reported for the whole year. The two can match, but they do not always match.

An $85,000 fortnightly-pay example

With no HELP, a full-year Australian resident claiming the tax-free threshold and earning $85,000 over 26 equal pays has the following modelled result for 2026–27:

Regular pay only; the standard work deduction applies to the annual estimate, with no other deductions, income or Medicare variations.
Gross each fortnight$3,269.23
Income tax and Medicare withheld each pay$682.00
Total withholding across 26 pays$17,732.00
Modelled annual tax and Medicare liability$17,400.00
Difference$332.00

The $332.00 difference includes the standard deduction and pay-period rounding; it is a modelled comparison, not a promise of a refund. Your actual tax return can contain other amounts.

Check the right payroll settings

The 2026 withholding instrument uses separate scales for residency, the tax-free threshold and Medicare exemptions. Study-loan withholding is calculated separately under Schedule 8. A bonus or back payment needs the applicable additional-payment rules, including Schedule 5, rather than treating every payment as ordinary salary.

Source: Taxation Administration (Withholding Schedules) Instrument 2026, Schedules 1, 5 and 8.

Compare like with like

In Payko, choose Payslip estimate for this regular-pay example. Choose Annual tax estimate for a whole-year comparison. For changing earnings, the year-to-date planner lets you enter earnings and withholding so far, plus expected future pay.

What this example does not settle

Income outside the model, deductions, family circumstances, tax offsets, loan balances and payroll adjustments can affect the final comparison. The model excludes extra-pay-year withholding additions and bonus withholding. Annual Medicare low-income thresholds are provisional. See calculation assumptions and dated sources.