Audit testing is the set of procedures an independent auditor uses to gather evidence about a company’s financial records and decide whether those records are free of material misstatement. The work has two sides: checking the internal systems that produce the numbers, and checking the numbers themselves. Together they support the auditor’s formal opinion on the financial statements, which is what shareholders, lenders, and regulators actually rely on.
Why Auditors Test in the First Place
Every audit comes down to one question: do the financial statements contain errors large enough to change someone’s decision? Those errors — called material misstatements — can come from honest math mistakes, flawed accounting judgments, or outright fraud. Under generally accepted auditing standards, auditors must obtain enough appropriate evidence to support their conclusions about a company’s financial position, and testing is the mechanism for gathering it.
The process begins with risk assessment. Auditors weigh two things before picking their tests: how likely a given account or transaction type is to contain a misstatement on its own (inherent risk), and how effective the company’s internal safeguards are at catching that misstatement before it reaches the financial statements (control risk). The combination determines how much direct testing is needed. High risk on either front means more testing. Low risk on both means the auditor can lean more heavily on the controls already in place.
One boundary worth flagging up front. Internal audits — the ongoing reviews performed by a company’s own staff for management and the audit committee — use similar techniques but exist for operational improvement. When people talk about “the audit,” they almost always mean the external audit performed by an independent CPA firm, and the testing described below applies to that work.
Control Testing
Auditors start with the internal systems a company uses to prevent mistakes before they reach the permanent records. Controls range from simple procedures like requiring two signatures on large checks to software restrictions that prevent one employee from both recording a sale and handling the cash. That separation of duties is one of the most basic controls, because it stops a single person from stealing money and hiding the theft in the ledger.
Testing asks two questions: was the control designed correctly, and did it actually work throughout the year? An auditor might pull a sample of purchase orders to confirm that every one above a set dollar threshold was approved by a supervisor, or check access logs to verify that only authorized users modified accounting records. If controls prove reliable, the auditor can scale back the direct number-checking required later. Weak controls have the opposite effect. They signal a higher risk of hidden errors and demand a deeper dive into the records themselves.
When controls fall short, the finding gets classified. A significant deficiency is serious enough to warrant the attention of those overseeing financial reporting. A material weakness is worse: it means there is a reasonable possibility that a material misstatement will not be caught in time.1PCAOB Public Company Accounting Oversight Board. AS 1305 Communications About Control Deficiencies in an Audit of Financial Statements When an auditor identifies a material weakness, the company cannot receive a clean opinion on its internal controls, and the finding becomes public in the annual filing.
Substantive Testing
Substantive testing shifts focus from the systems that create data to the data itself. These tests target specific dollar amounts and disclosures in the financial statements. Auditors work at the assertion level, meaning they challenge specific claims embedded in every line item: does this asset actually exist? Is this liability recorded at the right amount? Were these revenues recognized in the correct period?
Even when controls look strong, some substantive testing is always required. Controls reduce the volume of direct checking needed, but they never eliminate it. A company might have perfect procedures for recording sales, and the auditor still needs to confirm that the receivables on the balance sheet correspond to real customers who actually owe those amounts.
Tests of Details
Tests of details examine individual transactions or balances directly. That includes pulling a sample of invoices to verify that recorded expenses match the supporting documents, confirming bank balances with the bank, or recalculating depreciation schedules line by line. This is the granular, transaction-by-transaction work most people picture when they think about auditing.
Substantive Analytical Procedures
Analytical procedures take a wider view. Instead of checking individual entries, the auditor compares account balances or financial trends against expectations developed from prior-year data, industry benchmarks, or known relationships between accounts. If total salary expense jumped 40 percent while headcount stayed flat, that discrepancy demands an explanation. Analytical procedures can sometimes catch problems that transaction-level testing would miss entirely. An unauthorized payroll entry might not look suspicious on its own but would show up when aggregate payroll is compared to staffing records.2PCAOB. AU Section 329 Substantive Analytical Procedures
Auditors typically use both approaches together. Some assertions respond well to analytical procedures — payroll, rent, utilities, and other accounts with predictable patterns. Others require tests of details, including inventory existence, complex financial instruments, and one-time transactions.2PCAOB. AU Section 329 Substantive Analytical Procedures Getting this mix wrong is one of the most common audit deficiencies the PCAOB flags during its inspections.
Techniques for Gathering Evidence
Whether the goal is testing controls or testing balances, auditors draw on a standard toolkit. Each procedure produces a different kind of evidence with a different level of reliability.
- Inspection. Examining documents and records — contracts, invoices, bank statements, board minutes — to find support for recorded transactions. Strength depends on the source: a third-party bank statement is more persuasive than an internally generated memo.
- Observation. Watching a process in real time, such as visiting a warehouse to observe the physical inventory count. Observation provides strong evidence for that moment but says nothing about what happens when the auditor is not present.
- External confirmation. Sending requests directly to third parties — banks, customers, lenders — asking them to verify account balances or terms. One of the most reliable techniques, because the information comes from outside the company being audited.
- Recalculation. Re-performing math independently, whether that means recomputing depreciation, verifying interest accruals, or checking tax provisions. If the auditor’s number matches the company’s number, that is strong evidence of accuracy.
- Inquiry. Asking management and employees questions about processes, unusual transactions, or known issues. Inquiry alone is rarely sufficient to support a conclusion but often points the auditor toward areas that need deeper testing.
- Analytical procedures. Comparing financial data against expectations, as described above, to identify anomalies that warrant further investigation.
Modern audits increasingly rely on data analytics software that can scan entire transaction populations rather than small samples. These tools flag statistical outliers, duplicate payments, transactions posted outside business hours, or entries just below approval thresholds. The software does not replace professional judgment, but it lets auditors focus their manual effort on the entries most likely to contain errors.
Sampling and Materiality
No auditor checks every transaction. The volume at most companies makes that impractical, and auditing standards do not require it. Instead, auditors use sampling: selecting a representative subset of transactions or balances and testing those to draw conclusions about the full population. Sampling can be statistical (using probability theory to quantify the risk of an undetected error) or non-statistical (relying on professional judgment to select items). Both approaches require a clear link between the sample size and the risk the auditor is willing to accept.
The concept that drives sample size is materiality — the dollar threshold above which an error would reasonably influence an investor’s decision. There is no single formula. Common benchmarks include 5 to 10 percent of pretax income, 0.5 to 1 percent of total revenue, and 1 to 2 percent of total assets, with the choice depending on circumstances. The SEC has noted that while a 5 percent threshold provides a reasonable starting point for preliminary assessment, it should never be applied mechanically. Qualitative factors like whether the misstatement involves fraud, masks a change in earnings trends, or affects regulatory compliance can make an otherwise small error material.3U.S. Securities & Exchange Commission. SEC Staff Accounting Bulletin No. 99 Materiality
Auditors also set a second, lower threshold called performance materiality. The logic is straightforward. If overall materiality for the financial statements is $500,000, and the auditor only tests to that exact level, there is a risk that several individually immaterial misstatements could add up to a material total. Performance materiality — typically set between 50 and 85 percent of overall materiality — provides a cushion against that aggregation risk. The exact percentage is a judgment call influenced by factors like the company’s history of misstatements and the strength of its controls.
What the Testing Produces: The Audit Opinion
All of the testing leads to one deliverable: the auditor’s formal conclusion about whether the financial statements are presented fairly. Four types of opinions exist, and the distinction matters enormously to investors and regulators.
- Unqualified (clean) opinion. The financial statements present fairly, in all material respects, the company’s financial position in conformity with generally accepted accounting principles. This is what every company wants and what most receive.4Public Company Accounting Oversight Board. AS 3101 The Auditor’s Report on an Audit of Financial Statements When the Auditor Expresses an Unqualified Opinion
- Qualified opinion. The financials are fairly presented except for a specific issue. The auditor uses the phrase “except for” to describe the problem, whether that is a scope limitation that prevented testing of one area or a departure from GAAP that is material but not pervasive.5PCAOB Public Company Accounting Oversight Board. AS 3105 Departures From Unqualified Opinions and Other Reporting Circumstances
- Adverse opinion. The financial statements do not present fairly. The auditor concluded that the departures from GAAP are so material and pervasive that the financials as a whole are unreliable.5PCAOB Public Company Accounting Oversight Board. AS 3105 Departures From Unqualified Opinions and Other Reporting Circumstances
- Disclaimer of opinion. The auditor cannot form an opinion at all, usually because the company restricted the audit’s scope so severely that there was not enough evidence to reach any conclusion. A disclaimer is not appropriate when the auditor has evidence of material GAAP departures. That calls for a qualified or adverse opinion instead.5PCAOB Public Company Accounting Oversight Board. AS 3105 Departures From Unqualified Opinions and Other Reporting Circumstances
A qualified or adverse opinion on a public company’s financials can trigger stock price declines, covenant violations on debt agreements, and increased regulatory scrutiny. A disclaimer can be even more damaging, because it signals that the company may have actively obstructed the audit process.
Independence and Oversight
None of the testing means anything if the auditor has a financial or personal stake in the outcome. Independence is the foundation that makes audit opinions credible, and both federal law and professional standards enforce it through specific structural requirements.
Under auditor independence rules enforced by the PCAOB, the lead audit partner and the concurring review partner must rotate off a client engagement after five consecutive years in either role and then sit out for five years before returning.6Public Company Accounting Oversight Board. PCAOB Sanctions Blue and Co LLC for Auditor Independence and Quality Control Violations Rotation prevents the kind of long-term relationships where an auditor starts identifying with the client’s interests rather than the public’s.
The Sarbanes-Oxley Act also bars audit firms from providing certain non-audit services to their audit clients. The prohibited list includes bookkeeping, financial information systems design, appraisal and valuation services, actuarial services, management functions, human resources services, and legal services, among others.7U.S. Securities and Exchange Commission. Revision of the Commissions Auditor Independence Requirements An auditor cannot objectively evaluate financial statements they helped prepare, or controls they helped design.
The PCAOB backs these rules with inspections. In 2024, PCAOB inspectors found deficiencies in 39 percent of all audits reviewed, down from 46 percent in 2023 but still a rate that shows how often audit testing falls short in practice. The Big Four firms fared better at a 20 percent deficiency rate, while smaller firms inspected on a triennial cycle had rates as high as 61 percent.8Public Company Accounting Oversight Board. PCAOB Posts Report Detailing Significant Improvements Across Largest Firms Alongside Inspection Results in Record Time The quality of an audit depends heavily on who performed it, and those numbers set the realistic ceiling on how much reasonable assurance the process actually delivers.