Internal Control Evaluation: Scoping, Testing, and Deficiencies

An internal control evaluation is a structured review that tests whether your controls over financial reporting are properly designed and actually working in practice. The work follows a top-down, risk-based sequence: define what’s in scope based on where a material misstatement could occur, test the relevant controls for both design and operating effectiveness, then classify and report any deficiencies so they can be fixed. The steps below walk through how to do that, what evidence to gather, and how to handle what you find.

Who Has to Perform One

Section 404(a) of the Sarbanes-Oxley Act requires management of every public company to include an internal control report in its annual filing containing management’s own assessment of whether controls over financial reporting are effective as of fiscal year-end.1GovInfo. Sarbanes-Oxley Act of 2002 Section 404(b) adds a second obligation for larger filers: the independent auditor must separately attest to and report on management’s assessment.2Securities and Exchange Commission. Study of the Sarbanes-Oxley Act Section 404 Companies that are neither accelerated nor large accelerated filers, and emerging growth companies, are exempt from 404(b) but still must perform management’s own assessment under 404(a).

Private companies are not subject to SOX, but many run evaluations voluntarily when preparing for an IPO, seeking financing, or responding to board expectations. Federal executive branch agencies follow the GAO’s Green Book, and nonprofits receiving federal funds may adopt it voluntarily.3U.S. GAO. Standards for Internal Control in the Federal Government

The Framework You Evaluate Against

The COSO Internal Control—Integrated Framework, refreshed in 2013, is the most widely accepted benchmark for evaluating internal controls over financial reporting.4Committee of Sponsoring Organizations of the Treadway Commission (COSO). Internal Control It organizes internal control into five components, each supported by underlying principles. All five must be present and functioning for the system to be considered effective.

The five components are the Control Environment (the tone, ethics, structure, and competence the organization sets from the top), Risk Assessment (identifying and analyzing what could go wrong, including fraud risk), Control Activities (the specific approvals, reconciliations, segregation of duties, and technology controls that address identified risks), Information and Communication (getting reliable information to the right people at the right time), and Monitoring Activities (ongoing and periodic evaluations that confirm the other four are working). The formal evaluation you’re running is itself a monitoring activity.

Beneath the five components sit 17 principles that give you a concrete checklist. The Control Environment principles cover board independence, ethical values, and accountability for control responsibilities. Risk Assessment includes a principle specifically addressing fraud risk. Control Activities has a principle dedicated to technology controls. A gap in any single principle can mean the related component is defective, so evaluators use the 17 as their working reference.

Step 1: Scope the Evaluation Top-Down

The SEC’s interpretive guidance describes a top-down, risk-based approach as the most efficient way to structure the work: start at the financial statement level, identify where a material misstatement is most likely, then work downward to the specific controls that address those risks.5U.S. Securities and Exchange Commission. Commission Guidance Regarding Management’s Report on Internal Control Over Financial Reporting This prevents the common mistake of testing every control with equal intensity regardless of risk.

Start with Entity-Level Controls

Entity-level controls are policies, governance structures, and oversight mechanisms that operate across the whole organization rather than inside a single process. Examples include the board’s oversight of financial reporting, the code of conduct, the risk assessment process led by senior finance leadership, and the company’s monitoring systems.6Public Company Accounting Oversight Board. PCAOB Auditing Standard 2201 – An Audit of Internal Control Over Financial Reporting They vary in precision. A code of conduct creates an environment where process-level controls are more likely to work but doesn’t directly catch misstatements. A detailed budget-to-actual variance analysis run by management may operate precisely enough to address a specific risk without any additional lower-level testing.

Evaluating entity-level controls first lets you calibrate how much process-level testing you actually need. If certain entity-level controls address a financial reporting risk on their own with sufficient precision, you may not need to test additional controls for that risk. Weak entity-level controls should push you the other way and increase process-level testing.

Identify Significant Accounts and Map to Assertions

After entity-level work, identify the financial statement accounts and disclosures where a material misstatement is reasonably possible. Revenue recognition, inventory valuation, and complex estimates typically land in scope because of their inherent complexity and the judgment involved. Use quantitative materiality thresholds to guide these decisions, and apply qualitative judgment too. An account small in dollar terms can still matter if a misstatement there would change an investor’s perception.

Each in-scope control should map to one or more financial statement assertions. Auditing standards define these in categories that include existence (recorded assets and liabilities actually exist), completeness (all transactions that should be recorded are recorded), and valuation (amounts are stated at appropriate figures).7Public Company Accounting Oversight Board. Auditing Standard No. 15 – Audit Evidence A three-way match in the purchasing cycle, for instance, directly addresses the existence assertion for accounts payable: if the purchase order, receiving report, and vendor invoice don’t agree, you’ve caught a payable that shouldn’t be recorded or an amount that’s wrong.

Document upfront which business units, IT systems, and specific control owners are included. This prevents scope creep once testing starts and keeps everyone clear on what is and isn’t being evaluated.

Step 2: Test Design and Operating Effectiveness

The evaluation tests two distinct things about every in-scope control. Design effectiveness asks whether the control, if operated properly, would prevent or detect a material misstatement. Operating effectiveness asks whether the control is actually being performed as designed, consistently, by the people responsible for it. A control can be well designed on paper and completely ineffective in practice if the person responsible skips the review or rubber-stamps approvals without looking at the data. Testing must cover both.

Start Each Process with a Walkthrough

Before testing individual controls, walk through each significant process. A walkthrough traces a single transaction from its origin through every processing step to the point where it hits the general ledger. PCAOB standards describe four objectives: understanding how transactions flow through the process, identifying points where a misstatement could arise, identifying the controls management has implemented at those points, and identifying controls over unauthorized use of company assets.6Public Company Accounting Oversight Board. PCAOB Auditing Standard 2201 – An Audit of Internal Control Over Financial Reporting

Ask the people who actually perform the work how they handle the transaction, what they check, and what happens when something looks off. These conversations often reveal more than any documentation review. A control that looks robust in a policy manual may be performed inconsistently, or the person performing it may not fully understand what they’re looking for. Walkthroughs combine questioning, observation, document review, and re-performing key steps.

Four Evidence-Gathering Methods

Control testing uses four core methods, each with a different level of persuasiveness:

  • Inquiry. Asking personnel how they perform their control responsibilities. This is the least persuasive method on its own and must always be backed up by at least one other method. People will tell you they perform a control even when evidence suggests otherwise.
  • Observation. Watching someone perform the control in real time, such as a physical inventory count or a system access review.
  • Inspection. Examining documents or system records that prove the control was performed, such as approval signatures, reconciliation reports, or exception logs showing a supervisor’s review.
  • Reperformance. Independently executing the control yourself to confirm the results. Recalculating depreciation expense or re-executing a bank reconciliation gives you the most direct evidence of whether the control works.

Reperformance is the gold standard because you’re proving the math yourself rather than relying on a signature on a page. Most evaluations use a combination of all four, with the mix driven by the risk level of the control being tested.

Test IT General Controls Alongside Application Controls

Nearly every financial control today depends on technology, so IT general controls (ITGCs) are the foundation that makes application-level controls trustworthy. PCAOB standards reference controls over program changes, access to programs, and computer operations as categories that must be effective for automated application controls to be reliable.6Public Company Accounting Oversight Board. PCAOB Auditing Standard 2201 – An Audit of Internal Control Over Financial Reporting In practical terms, ITGC testing covers whether the right people have the right access (and only the right access) to financial systems, whether changes to those systems go through a controlled approval and testing process before deployment, and whether backups and system operations run reliably.

Strong ITGCs let you do less repetitive work on automated controls. If you can verify that an automated control hasn’t changed since it was last tested, you may be able to rely on it without repeating full operational testing every period. Weak ITGCs undermine confidence in every automated control that depends on them.

Choose a Defensible Sample Size

You can’t test every instance of a control that runs daily across hundreds of transactions. Sampling lets you draw conclusions about the full population from a representative subset. The appropriate sample size depends on how often the control operates, the risk of the account or assertion it addresses, whether the control is manual or automated, and the competence and consistency of the person performing it.6Public Company Accounting Oversight Board. PCAOB Auditing Standard 2201 – An Audit of Internal Control Over Financial Reporting

A daily manual control needs a significantly larger sample than a quarterly management review because there are more opportunities for failure and more transactions at risk. Automated controls with effective ITGCs typically require smaller samples because the system performs the control the same way every time. Whatever method you use, apply it consistently and document both your rationale for the sample size and how you selected the items.

Step 3: Classify Deficiencies by Severity

When testing reveals a control that didn’t work as designed, the next step is determining how serious it is. Not every failure is equally dangerous, and the classification drives what happens next.

  • Control deficiency. The control’s design or operation doesn’t allow the people responsible for it to catch misstatements while performing their normal work. An example is a required review that happens but isn’t documented, making it impossible to verify later. The vast majority of deficiencies identified in practice fall into this category.8U.S. Securities and Exchange Commission. Sarbanes-Oxley Section 404 Costs and Remediation of Deficiencies
  • Significant deficiency. A deficiency, or a combination of deficiencies, less severe than a material weakness but important enough to merit attention from those overseeing financial reporting. Multiple smaller deficiencies that individually seem manageable can aggregate into a significant deficiency when they affect the same process or assertion.6Public Company Accounting Oversight Board. PCAOB Auditing Standard 2201 – An Audit of Internal Control Over Financial Reporting
  • Material weakness. A deficiency, or combination of deficiencies, where there is a reasonable possibility that a material misstatement will not be prevented or detected on a timely basis. “Reasonable possibility” means the likelihood is either reasonably possible or probable, and the bar is lower than many people assume.6Public Company Accounting Oversight Board. PCAOB Auditing Standard 2201 – An Audit of Internal Control Over Financial Reporting

Classification requires judgment, not a formula. You’re estimating both the likelihood that the control failure could lead to a misstatement and the magnitude of that potential misstatement. A small likelihood of a massive misstatement can be just as serious as a high likelihood of a moderate one.

For public companies, the stakes escalate sharply at the material weakness level. SEC rules require that management’s annual assessment disclose any material weaknesses identified, and management cannot conclude that internal controls are effective if even one material weakness exists.9U.S. Securities and Exchange Commission. Management’s Report on Internal Control Over Financial Reporting Companies subject to auditor attestation under Section 404(b) also have the auditor publicly report material weaknesses that exist as of the assessment date.8U.S. Securities and Exchange Commission. Sarbanes-Oxley Section 404 Costs and Remediation of Deficiencies

Step 4: Report, Remediate, and Re-Test

The evaluation report should describe each deficiency clearly: what the control was supposed to do, what actually happened, and why. Identifying the root cause matters more than most teams realize. A failed reconciliation might trace back to a training gap, an understaffed accounting team, a poorly designed approval workflow, or a system that doesn’t flag exceptions properly. Each root cause points to a different fix, and getting this wrong means the problem recurs in the next evaluation cycle. Quantify the potential financial exposure where possible, because abstract descriptions of risk don’t generate the urgency that dollar figures do.

A remediation plan needs three elements: a clear owner (a specific person, not a department), concrete corrective actions (new automated controls, revised procedures, targeted training), and a realistic deadline. Vague commitments to “improve the process” accomplish nothing.

After enough time has passed for the corrective actions to take effect, re-test the remediated controls. Successful re-testing closes the loop and provides evidence that the risk has been addressed. Failed re-testing means the root cause analysis was wrong or the fix was inadequate, and the work goes back to the start. Many organizations lose discipline here: the initial evaluation gets done rigorously, but follow-through on remediation quietly drops off the priority list.

Beyond the formal cycle, embed monitoring into daily operations. Automated system checks, supervisory reviews built into transaction approval workflows, and ongoing reconciliation procedures generate real-time signals about whether controls are holding up, and they catch minor deviations before they compound into reportable deficiencies.