Audit Evidence: Reliability, Sufficiency, and Practical Methods

Learning objectives

By the end of this chapter, you should be able to:

• Explain what audit evidence is and distinguishrelevancefromreliabilitywhen evaluating it.
• Judge how much evidence is required in different situations, linking the extent of work to risk and materiality.
• Select practical evidence-gathering methods (for example inspection, observation, recalculation, re-performance, confirmation, enquiry and analytical procedures) and match them to common assertions.
• Deal appropriately with conflicting information and understand when additional procedures are required.
• Document audit work so that an experienced reviewer can understand what was done, what was found, and how conclusions were reached.
• Recognise common evidence pitfalls, including bias, weak corroboration and over-reliance on enquiry.

Overview & key concepts

Audit evidence sits behind every audit conclusion. It consists of information obtained and evaluated to support whether the financial statements are free from material misstatement. The auditor responds to the assessed risk of material misstatement by designing further audit procedures that reduce detection risk. When those procedures are properly performed and evaluated, overall audit risk is reduced to an acceptably low level, supported by sufficient appropriate evidence.

Evidence work links directly to the financial statements. If evidence supports (or challenges) a balance or transaction, it supports (or challenges) the related elements:

• Assets(for example receivables, inventory, property)
• Liabilities(for example payables, loans, provisions, deferred income)
• Equity(for example share capital and retained earnings)
• Income and expenses(for example revenue, cost of sales, operating expenses)

Audit evidence

Audit evidence is the information used to form audit conclusions. It is obtained from:

• Accounting records(ledgers, journals, invoices, contracts, reconciliations, management reports)
• Other sources(third parties, external databases, observations, confirmations, physical inspection, recalculation)

Two questions drive evidence evaluation:

1. Is it relevant?
2. Does it address the specific assertion and period being tested?
3. Is it reliable?
4. How much confidence can be placed in it, given where it came from and how it was obtained?

Sufficiency and appropriateness

Appropriateness (quality)

Appropriateness is about quality, mainly:

• Relevanceto the assertion being tested
• Reliabilityof the source and method

Better quality evidence can reduce the extent of further work, but it does not remove the need for enough coverage over the population.

Sufficiency (quantity)

Sufficiency is about quantity: how much evidence is needed to support a conclusion. The amount of work is influenced by:

• therisk of material misstatementand the specific audit response required
• materiality and tolerable misstatementfor the balance or class of transactions
• sampling risk(the chance the sample does not represent the population) andpopulation characteristics(size, variability, stratification, expected error)
• theexpected misstatement rate(based on prior experience, walkthroughs, interim results, or preliminary testing)
• theeffectiveness of controls, where the audit approach plans to rely on them (strong tested controls can reduce the extent of substantive work; weak controls usually increase it)

Quantity alone never compensates for poor-quality evidence. If evidence is weak or inconsistent, the response is usually to obtain stronger evidence and/or extend testing.

How auditors think: double-entry logic

Understanding double-entry is essential because misstatements rarely sit in isolation. Errors and manipulation typically affect at least two places.

Examples:

• Overstated revenue may also overstate receivables (or understate deferred income) and overstate profit and equity.
• Understated payables may understate expenses or cost of sales, inflating profit and equity.
• Inventory errors affect assets and profit through cost of sales.

This logic guides testing direction:

• Vouching(ledger → source) is powerful for testingoccurrence/existence.
• Tracing(source → ledger) is powerful for testingcompleteness.

Reliability of audit evidence

How to judge reliability in practice

Rather than treating sources as automatically “strong” or “weak,” assess reliability using four practical questions:

• Independence:How far is the information from management influence (for example third-party data vs internally prepared schedules)?
• Control and custody:Could the client intercept, edit, or curate what you receive (for example replies routed through management, emailed PDFs, screenshots)?
• Integrity of the item:Is it traceable to an underlying record with a clear audit trail, or is it an extract/copy that could be altered without detection?
• System strength:If evidence is produced by the entity’s systems, are access rights, change controls, and review controls strong enough to trust the output?

Reliability generally increases when evidence is independent, received directly by the auditor, traceable to underlying records, and generated in a controlled environment. Auditors still need to remain alert to error, bias, and modern threats such as spoofed emails and fabricated documents.

Corroboration

Corroboration strengthens a conclusion by combining evidence from different sources or different procedures. It reduces the risk of relying on one narrow indicator.

Examples:

• A receivable confirmation supported bysubsequent cash receiptsandsales/dispatch documentation.
• Inventory count observations supported bycut-off testing,price testing, andnet realisable value checks.
• Payroll expense supported byre-performance,authorised change documentation, andbank payment evidence.

Professional scepticism

Professional scepticism must be applied throughout planning and performance, not only when something looks wrong. It is particularly important when dealing with:

• contradictory evidenceand exceptions
• management estimatesand other judgemental areas
• related party transactionsand unusual relationships
• fraud risks, including the possibility of management override of controls

Scepticism is disciplined judgement: asking whether the evidence is complete, consistent, and plausible, and expanding work where it is not.

Contradictory evidence

Contradictory evidence occurs when information does not fit together. It must be resolved, not ignored. Responses commonly include:

• extending tests and obtaining alternative evidence
• investigating timing differences and cut-off errors
• checking for disputes, returns, credit notes, and post-year-end adjustments
• reassessing whether one exception indicates a broader control or fraud issue

A conclusion is not supported if significant contradictions remain unresolved.

Management representations

What they add—and what they don’t

Written representations are management’s formal confirmation of responsibilities and certain judgements. They can help close gaps where evidence is indirect (for example confirming that all known litigation has been disclosed), but they do not replace documentary or independent support.

If management refuses to provide representations, or if representations are unreliable or conflict with other findings, this is a serious limitation. The auditor must redesign procedures and consider the impact on the report, including whether a modified opinion is required depending on significance and pervasiveness.

External confirmation

External confirmation obtains information directly from an independent third party (for example banks, customers, suppliers). It can provide strong evidence for existence, rights and obligations, and sometimes valuation.

Key practical points:

• The auditor must control the process (selection, addressing, sending, and receiving replies).
• Non-responsesrequire follow-up and, where necessary, alternative procedures (for example subsequent receipts for receivables, supplier statements and post-year-end payments for payables).
• Reliability threats includeinterception or management involvement, incorrect contact details, andemail fraud/spoofing. Replies should be assessed for authenticity and authority (who responded, from where, and whether the response makes sense in context).
• Exceptions and disputed responses must be investigated and resolved.

Analytical procedures

Analytical procedures evaluate financial information through relationships such as trends, ratios, comparisons, and reasonableness models. They are used in planning and can also provide substantive evidence when relationships are predictable and data is reliable. Where judgement is high or relationships are unstable, analytical work usually needs support from tests of details.

Tests of controls and substantive procedures

Audit procedures are often described as:

• Tests of controls:to evaluate whether controls operated effectively during the period (supporting a plan to rely on controls).
• Substantive procedures:to detect material misstatements at the assertion level, including analytical procedures and tests of details.

How evidence choices change when controls are strong or weak

Where controls are designed well and have been tested as operating effectively, the audit response may place more emphasis on control testing and reduce the extent of substantive testing (while still performing sufficient substantive work overall). Where controls are weak, the audit approach typically shifts towards more extensive substantive procedures, larger sample sizes, and more direct evidence (for example confirmations, external documents, re-performance).

Core theory and frameworks

Assertions and evidence matching

Evidence is strongest when procedures are chosen to match the assertion under test. Assertions are framed slightly differently depending on what is being tested:

• Classes of transactions and events(for example revenue, purchases, payroll)
• Account balances(for example receivables, inventory, payables)
• Presentation and disclosures(for example classification, completeness of disclosures)

Common assertions include:

• Existence
• Completeness
• Accuracy
• Cut-off
• Valuation
• Rights and obligations
• Presentation and disclosure

A single procedure rarely covers all assertions; effective work usually combines procedures to provide corroboration.

Evidence and the accounting equation

The accounting equation (Assets = Liabilities + Equity) is a useful sense-check for audit findings. When evidence suggests a misstatement, consider:

• which balances are overstated/understated
• what double-entry would have been recorded (or omitted)
• whether profit is affected (and therefore retained earnings)
• whether the issue is measurement, classification, or disclosure

Worked example

Narrative scenario

ABC Ltd is a retailer preparing financial statements for the year ended 31 December 2025. The draft results include:

• Revenue:£1,675,000
• Reported gross margin:21.8%
• Capital expenditure during the year:£76,000
• Tax rate used by management in planning calculations:24.3%
• Discount rate used by management for certain long-term estimates:6.7%

During the audit, the following areas are identified as requiring focused evidence:

1. Sales revenue recorded close to year end
2. Inventory quantities and valuation at year end
3. Trade receivables and payables balances
4. Management’s estimate of warranty provisions
5. Documentation of procedures and conclusions

Required

• Evaluate the sufficiency and appropriateness of evidence for year-end revenue.
• Verify inventory existence and valuation at year end.
• Confirm the accuracy of receivables and payables.
• Assess the reliability of evidence supporting the warranty provision.
• Summarise and document procedures and conclusions.

Solution

1) Sales revenue recognition (year-end)

Risk focus: revenue cut-off and validity close to year end; risk of revenue inflation via early recognition or unsupported manual entries.
Key assertions: occurrence, accuracy, cut-off, and (where relevant) classification between revenue and deferred income.

Best procedure mix:

• Cut-off testing (directional):select sales recorded just before year end and match to dispatch/delivery evidence; select dispatches just before year end and ensure the related sales are recorded in the correct period.
• Inspect post-year-end credit notes/returns:investigate whether these indicate sales recorded prematurely or incorrectly.
• Confirm selected customer balances or specific invoiceswhere risk is high (large, unusual, or disputed items).
• Analytical expectation using gross margin:
  • Expected gross profit ≈ £1,675,000 × 21.8% =£365,150
  • Expected cost of sales ≈ £1,675,000 − £365,150 =£1,309,850
  • Use this as a reasonableness check and investigate unusual movements by product line, location, or week.

Financial statement impact (conceptual):
Overstated revenue usually overstates profit and equity, and may overstate receivables or cash. If invoicing occurs before performance, amounts may be more appropriately shown as deferred income rather than revenue.

Conclusion:
Revenue is supported when cut-off testing, returns/credit notes review, and (where appropriate) confirmations provide consistent evidence that recorded sales belong in the period and represent genuine transactions.

2) Inventory verification (year-end)

Risk focus: overstatement of inventory quantities and values, and cut-off errors affecting both inventory and cost of sales.
Key assertions: existence, completeness, valuation, cut-off.

Best procedure mix:

• Attend and observe the inventory count,perform test counts, and assess count controls and count instructions.
• Reconcile count results to inventory recordsand investigate differences.
• Cut-off testing:inspect goods received notes and dispatch records around year end to ensure purchases and sales are recorded in the correct period.
• Cost testing:agree unit costs to supplier invoices and verify that costing is applied consistently.
• NRV testing:compare selling prices after year end (less selling costs) to carrying amounts, focusing on slow-moving, obsolete, or damaged lines.

Financial statement impact (conceptual):
Inventory errors affect assets and profit via cost of sales. Overstated closing inventory typically overstates profit and equity.

Conclusion:
Inventory is supported when quantity evidence (count) and valuation evidence (cost/NRV) are both obtained and are consistent with cut-off testing results.

3) Trade receivables and trade payables

Trade receivables

Risk focus: existence (fictitious debtors) and valuation (uncollectible balances).
Key assertions: existence, valuation, rights.

Best procedure mix:

• Customer confirmationsfor a sample of balances (or specific invoices), controlling the process.
• Subsequent cash receipts testingto support existence and collectability.
• Aged receivables reviewand inspection of dispute correspondence.
• Impairment assessment:evaluate the reasonableness of expected credit losses by testing inputs and considering whether specific high-risk balances need additional provision.

Conclusion:
Receivables are supported when confirmations and/or subsequent receipts support existence, and the impairment approach is consistent with observable collection evidence.

Trade payables

Risk focus: understatement through unrecorded liabilities.
Key assertions: completeness, existence, accuracy.

Best procedure mix:

• Supplier statements reconciliationto the payables ledger.
• Search for unrecorded liabilities:review post-year-end payments, unmatched goods received notes, and unmatched supplier invoices.
• Confirm selected supplier balanceswhere reconciling items, disputes, or unusual movements exist.

Conclusion:
Payables are supported when completeness-focused procedures (supplier statements and unrecorded liabilities work) provide consistent evidence that liabilities are not understated.

4) Warranty provisions

Risk focus: management bias and uncertainty in estimates; completeness of obligations and measurement.
Key assertions: completeness, valuation.

Best procedure mix:

• Inspect warranty termsto confirm which sales create obligations and the coverage period.
• Re-perform the calculationand test the accuracy of underlying data (eligible sales volumes, claim rates, average repair/replace costs).
• Compare prior estimates to actual outcomesto assess estimation reliability.
• Assess current conditions(new products, known defects, supplier issues) that could change claim patterns.

Discounting (where relevant):
If expected cash outflows are spread over a long period and the impact is material, consider whether discounting is appropriate and whether the discount rate used is reasonable. For short settlement periods, discounting is unlikely to be significant.

Conclusion:
The provision is supported when the model is re-performed, inputs are validated, and the output is consistent with historical outcomes and current conditions, with discounting considered where relevant.

5) Working papers and clear sign-off

Your audit file should tell the story of the work without needing you in the room to explain it. For each test, record:

• what you were trying to prove (the risk and the assertion)
• what you tested (the population, the selection method, and the items chosen)
• what you did (steps performed, timing, and who did the work)
• what you saw (key evidence captured or referenced so it can be re-traced)
• what went wrong (exceptions, how they were investigated, and the final outcome)
• why the conclusion makes sense (how results link back to risk and materiality)

Good documentation is not just a checklist of actions; it shows that the work was targeted, sufficient for the risk, and properly evaluated.

Interpretation of the results

The procedures above show how evidence decisions depend on risk and controls:

• Year-end revenue typically needs strong cut-off work and corroboration from dispatch/delivery evidence and post-year-end returns.
• Inventory conclusions require quantity and valuation work, supported by cut-off testing.
• Receivables testing must address both existence and impairment; payables testing must prioritise completeness.
• Provisions require evidence over assumptions and inputs, with a clear evaluation of potential management bias.

Common pitfalls and misunderstandings

• Treating enquiry as evidence on its own:explanations guide work but rarely provide sufficient support without corroboration.
• Confusing relevance with reliability:evidence must address the assertion and be trustworthy.
• Ignoring completeness risks in liabilities:payables and accruals are commonly understated; testing must search beyond recorded items.
• Stopping after inventory count attendance:valuation and cut-off still require separate evidence.
• Over-trusting internal outputs:internal reports require control evaluation and corroboration.
• Not aggregating exceptions:individual small errors may point to a pattern when combined.
• Weak cut-off thinking:cut-off is directional and often requires testing from both sides of year end.
• Unresolved contradictions:conflicting evidence must trigger extra procedures and a reassessment of risk.
• Assuming confirmations always solve the issue:non-responses and disputed replies require alternative work.

Summary and further reading

Audit evidence supports conclusions reached from procedures designed to reduce detection risk to an acceptably low level. Appropriateness is about quality (relevance and reliability); sufficiency is about quantity and depends on risk, materiality/tolerable misstatement, sampling risk, expected misstatement rates, and whether effective controls can be relied on. Strong audits use corroboration, investigate contradictions, apply professional scepticism throughout, and document work so the file clearly supports the final conclusions.

For wider context, review materials on audit risk, internal controls, estimates and provisions, revenue cut-off, and analytical procedures.

FAQ

Why is external confirmation often stronger than internal evidence?

Because it is obtained from an independent party and is less exposed to management bias. However, confirmations must be controlled by the auditor and assessed for authenticity, authority of the respondent, and reliability threats such as interception or email spoofing.

How is the sufficiency of audit evidence determined?

By linking the planned work to the risk of material misstatement and materiality/tolerable misstatement, considering sampling risk and population characteristics, and evaluating expected error rates. Where controls are tested and reliable, the extent of substantive work may reduce; where controls are weak, substantive work usually increases.

What does professional scepticism change in practical terms?

It changes how the auditor responds to estimates, unusual items, contradictions, related party activity, and fraud risks. It drives a requirement to corroborate explanations and to expand procedures when evidence is inconsistent or overly dependent on management judgement.

What are common pitfalls when using analytical procedures?

They become weak when expectations are unclear, data quality is poor, or relationships are not stable. Analytical results must be supported by investigation of significant deviations and, where needed, by tests of details.

What should be done when evidence conflicts?

The conflict must be resolved through additional procedures and careful evaluation. If unresolved issues remain significant, the auditor must reconsider the risk assessment and the impact on the report.

Glossary

Audit evidence
Information used to support audit conclusions, obtained from accounting records and other sources.

Sufficiency
The amount of evidence needed to support a conclusion, influenced by risk of material misstatement, materiality/tolerable misstatement, sampling risk, expected misstatement, and control effectiveness where relying on controls.

Appropriateness
The quality of evidence, mainly its relevance to the assertion and its reliability.

Reliability
The degree of confidence that can be placed in evidence, influenced by independence, custody/control, traceability to underlying records, and the strength of systems and controls that generate the information.

Corroboration
Strengthening a conclusion by combining evidence from different sources or different procedures.

Professional scepticism
A questioning approach applied throughout planning and performance, especially for estimates, contradictions, related parties, and fraud risks.

Contradictory evidence
Information that conflicts with other evidence and must be investigated and resolved.

Management representations
Written confirmations from management that support certain matters but do not replace other evidence; refusal or unreliability can create a limitation that may affect the audit report.

External confirmation
Evidence obtained directly from a third party (for example banks, customers, suppliers), requiring control over the process and careful evaluation of authenticity and exceptions.

Analytical procedures
Evaluation of financial information by studying relationships and trends to identify inconsistencies or support conclusions.

Tests of controls
Procedures performed to evaluate whether controls operated effectively during the period.

Substantive procedures
Procedures designed to detect material misstatements, including analytical procedures and tests of details.

Tests of details
Substantive procedures that verify amounts and disclosures through inspection, recalculation, re-performance, and other direct checks of underlying records.