SA 530 Audit Sampling explained: this guide covers what it means, who it applies to, the step-by-step process, documents required, fees, due dates and penalties in India — so you can stay compliant with confidence and avoid costly mistakes.
SA 530 applies when an auditor decides to test less than every item in a population and draw a conclusion about the whole. It explains how the sample is designed, how big it must be, how items are chosen, and what to do with the errors found.
SA 530, as effective for audits of financial statements for periods beginning on or after 1 April 2009, covers both tests of controls and tests of details. ICAI may revise standards, so check icai.org for the current text. A sample is only as sound as the listing it comes from, so complete registers kept through books of accounts compliance matter.
The aim of sampling is a reasonable basis for conclusions about the whole population (paragraph 4). The auditor designs the sample for the purpose of the test, sets a size that brings sampling risk down to an acceptably low level, selects so that every item has a chance of selection, investigates each deviation or misstatement, projects misstatements found in tests of details to the population, and evaluates whether the sample gave a reasonable basis. Statistical and non-statistical approaches are both acceptable; the choice is a matter of judgement.
Scope and key terms (paragraphs 1-5)
The standard applies once the auditor has decided to use sampling. It complements SA 500, under which sampling is one of the means of selecting items for testing (paragraph 2).
| Term (paragraph 5) | Plain meaning |
|---|---|
| Audit sampling | Applying audit procedures to fewer than every item in a population, giving each sampling unit a chance of selection |
| Population | The complete set of data from which the sample is taken and about which the auditor wants to conclude |
| Sampling risk | The risk that the conclusion from the sample differs from the one reached if the whole population were tested |
| Non-sampling risk | Risk of a wrong conclusion for reasons unrelated to sampling, such as using an unsuitable procedure or misreading evidence (A1) |
| Anomaly | A misstatement or deviation demonstrably not representative of the population |
| Sampling unit | An individual item, such as an invoice, a cheque, a debtor's balance, or a rupee (A2) |
| Statistical sampling | Random selection plus probability theory to evaluate results, including measuring sampling risk; anything else is non-statistical |
| Stratification | Dividing a population into sub-groups with similar characteristics, often value |
| Tolerable misstatement | An amount set by the auditor for which assurance is sought that actual misstatement does not exceed it |
| Tolerable rate of deviation | The same idea, applied to control deviations |
Sampling risk has two sides. The auditor is mainly worried about concluding that controls are better than they are, or that no material misstatement exists when it does, because this affects the opinion. The opposite error leads only to extra work and affects efficiency (paragraph 5(c)). Tolerable misstatement is performance materiality (see SA 320) applied to a particular sampling procedure, and may be equal to or lower than it (A3).
Sample design (paragraph 6)
The auditor considers the purpose of the procedure and the characteristics of the population (A4-A9). Four points stand out:
- Define a deviation or misstatement precisely. In an existence test of receivables using confirmations, a customer payment made before the confirmation date but received by the client afterwards is not a misstatement, and a misposting between customer accounts does not change the total, though it may matter to fraud risk or the allowance for doubtful debts (A6).
- Make sure the population is complete. The auditor performs procedures to obtain evidence of this (A5).
- Estimate the expected deviation or misstatement. For controls, an unacceptably high expected deviation rate normally means the auditor does not test controls. For details, a high expected misstatement may mean examining every item or using a large sample (A7).
- Consider stratification. Appendix 1 explains how splitting a population, for example by value, improves efficiency.
The choice between statistical and non-statistical sampling is a matter of judgement, and sample size is not a valid way to tell them apart (A9).
Sample size (paragraph 7)
The size must reduce sampling risk to an acceptably low level. The less risk the auditor will accept, the larger the sample (A10). Size can come from a statistical formula or from judgement, and when circumstances are alike, the same factors push size in the same direction either way (A11). The appendices list factors, such as the level of assurance wanted, the tolerable rate or misstatement, and the expected rate or amount, in table form: more assurance, more sample; higher tolerance, smaller sample. For large populations the actual population size has a negligible effect.
Selecting the items (paragraph 8)
Each sampling unit must have a chance of selection. Statistical sampling gives each unit a known probability; non-statistical sampling uses judgement but must still aim at a representative sample that avoids bias (A12). Appendix 4 describes the methods:
| Method | In brief |
|---|---|
| Random selection | Units chosen through random number generation or similar |
| Systematic selection | A fixed interval is applied from a random start |
| Monetary unit sampling | Value-weighted selection where each rupee is a unit, so larger items are likelier to be chosen |
| Haphazard selection | Picking without a structured technique, but trying to avoid bias; not appropriate with statistical sampling |
| Block selection | A contiguous block of items; ordinarily not suitable for drawing conclusions about the whole population |
Performing the procedures (paragraphs 9-11)
The procedure is applied to each item selected (paragraph 9). If it is not applicable, for instance a properly cancelled cheque in a test of payment authorisation, a replacement is tested (paragraph 10, A14). If neither the procedure nor a suitable alternative can be applied, as with lost documents, the item is treated as a deviation in a test of controls or a misstatement in a test of details (paragraph 11, A15). An example of an alternative is examining later receipts when a debtor has not replied to a confirmation (A16); compare SA 505.
Deviations, anomalies and projection (paragraphs 12-14)
The auditor investigates the nature and cause of each deviation or misstatement and its effect on the purpose of the test and on other areas (paragraph 12). If many share a feature, such as a location or a period, the auditor may extend work to all items having it, and should also consider that they may indicate fraud (A17).
Treating something as an anomaly is for extremely rare cases, and the auditor needs a high degree of certainty, from additional procedures, that it does not affect the rest of the population (paragraph 13).
For tests of details, misstatements found are projected to the population (paragraph 14). Projection gives a view of the scale but may not be enough to fix an amount to be recorded (A18). An anomaly may be left out of the projection, but its effect if uncorrected is still considered (A19). For tests of controls, the sample deviation rate is itself the projected rate (A20).
Evaluating the results (paragraph 15)
An unexpectedly high deviation rate may raise the assessed risk unless other evidence supports the original assessment; an unexpectedly high misstatement may suggest the whole class or balance is materially misstated (A21). The projected misstatement plus any anomaly is the auditor's most reasonable estimate of misstatement in the population. If it exceeds tolerable misstatement, the sample does not give a reasonable basis, and the closer it is to tolerable misstatement, the more likely actual misstatement exceeds it (A22). In that case the auditor may ask management to investigate and adjust, or tailor further procedures, for instance by extending the sample, testing another control or modifying substantive procedures (A23).
Illustrative example
Kavya Retail Pvt Ltd is an invented company; all figures are illustrative. The auditor tests 2,400 purchase invoices for approval, using a systematic selection of 60 from a random start. Two have no approval signature, which the auditor treats as deviations; the auditor then investigates and finds both come from one branch in one month, so all invoices from that branch and month are examined. In a separate test of details on 120 customer balances (a book value of Rs 9 crore), sample errors project to Rs 22 lakh against a tolerable misstatement of Rs 18 lakh. The sample therefore does not give a reasonable basis, and the auditor asks management to investigate and extends the work.
Documentation and links
Sampling work is recorded under SA 230. The working papers should show the population, the definition of a deviation, the method and size, the items examined and the evaluation. See also our guide on tax audit procedures, SA 230 and SA 530 working papers. The text prints no modifications compared with the international standard.
Need help with sampling support?
Auditors sample from your registers, so clean, complete listings of invoices, vouchers and balances make the process quicker. TaxClue's books of accounts compliance support can help you keep ledgers and listings complete and reconciled before the audit starts.
Key takeaways
- Sampling must give a reasonable basis for conclusions about the whole population.
- The population must be complete and a deviation clearly defined before selection.
- Every item needs a chance of selection; the method must avoid bias.
- Misstatements in tests of details are projected; anomalies are very rare and need strong proof.
- If projected misstatement exceeds tolerable misstatement, more work is needed.
Read next
Disclaimer: Based on the Standards on Auditing and quality standards issued by the Institute of Chartered Accountants of India, in the versions named in the article, and ICAI's announcement of 31 March 2026 on SQM 1 and SQM 2, as consulted on 3 October 2026. ICAI revises standards from time to time; check the current text and effective dates on icai.org. This article is general information, not legal advice; check the official text before acting.