How the IRS Selects Returns for Audit

The IRS does not audit returns at random. A structured, multi-layered selection system identifies which of the more than 160 million individual returns filed annually (IRS Data Book) warrant examination. Understanding how that system operates — and where its thresholds fall — is central to interpreting any IRS correspondence or notice that initiates the audit process.

Definition and Scope

An IRS audit, formally called an examination, is the agency's review of a taxpayer's accounts and financial information to verify that income, expenses, and credits are reported correctly under the Internal Revenue Code. The authority to conduct examinations derives from IRC § 7602, which grants the IRS power to examine books, papers, records, and other data relevant to determining tax liability.

The examination function sits within the IRS Large Business and International (LB&I) division, the Small Business and Self-Employed (SB/SE) division, and the Wage and Investment (W&I) division — each targeting a distinct taxpayer population. The IRS audit process encompasses correspondence audits conducted by mail, office audits at an IRS facility, and field audits conducted at a taxpayer's home or place of business.

The overall audit rate for individual returns has declined substantially over the past decade. According to the IRS Data Book 2022, the IRS audited approximately 0.38% of individual income tax returns filed for tax year 2019 — fewer than 4 in every 1,000 returns. That aggregate figure, however, masks sharp variation by income level and return type.

How It Works

Return selection operates through three primary mechanisms, each applying different analytical tools and triggers.

1. Discriminant Information Function (DIF) Scoring

Every individual return processed through IRS systems receives a DIF score, a proprietary numeric value generated by a statistical model that compares reported figures against norms derived from past audits of similar returns. The IRS has never published the specific weights or formulas within the DIF model (Internal Revenue Manual, Part 4, Chapter 19). Returns with high DIF scores — indicating that reported deductions, credits, or income figures deviate significantly from statistical norms for that income bracket — are flagged for human review by classifiers, IRS employees who make the final selection decision.

2. Unreported Income Document Matching

The Automated Underreporter (AUR) program cross-references information returns — Forms W-2, 1099-NEC, 1099-INT, 1099-DIV, and similar documents filed by employers, financial institutions, and clients — against what the taxpayer reported. A mismatch between a 1099 amount and the figure on the return triggers a CP2000 notice, which proposes a tax adjustment. This program processes tens of millions of return comparisons annually and operates largely automatically before any human examiner becomes involved (IRS AUR Program, IRM Part 4, Chapter 19.2).

3. Specific Compliance Programs and Referrals

Beyond scoring and matching, returns may be selected through:

1. National Research Program (NRP) — periodic random sampling used to update DIF scoring norms and estimate the tax gap; selection under NRP is not based on suspected noncompliance.
2. Related return examinations — when a business or partnership is audited, the returns of owners, partners, or related entities may be pulled for consistency review.
3. Whistleblower referrals — under IRC § 7623, the IRS Whistleblower Office accepts information from third parties and may open examinations based on that information.
4. Criminal Investigation referrals — the IRS Criminal Investigation Division may refer a case for civil examination after a criminal inquiry.
5. Local compliance initiatives — LB&I and SB/SE publish compliance campaigns targeting specific industries or transaction types; returns falling within a campaign's scope face elevated selection probability.

The full overview of IRS functions and authorities is documented on the IRS authority resource index.

Common Scenarios

Three return profiles consistently generate elevated DIF scores or matching exceptions.

Self-Employment Income and Schedule C Filers

Returns reporting self-employment income via Schedule C draw disproportionate scrutiny because business expenses are self-reported without employer verification. The IRS audit rate for Schedule C filers reporting gross receipts above $100,000 runs materially higher than the 0.38% aggregate rate for all individual returns. Large home office deductions, vehicle expense claims at or near 100% business use, and meals and entertainment deductions at elevated percentages relative to reported income all contribute to higher DIF scores. Self-employment tax obligations compound the complexity, since SE filers must also reconcile net earnings for SECA purposes.

High-Income Returns

The IRS Data Book 2022 shows that returns reporting total positive income above $10 million were audited at a rate of approximately 8.7% — more than 22 times the overall average. High-income returns involve more line items, more schedules, and more opportunities for mathematical or substantive discrepancies, which increases both DIF scores and the likelihood of referral through related-party examination.

Earned Income Tax Credit Claims

Earned Income Tax Credit (EITC) claims generate a separate targeting stream due to historically elevated improper payment rates. The Office of Management and Budget reported an EITC improper payment rate of approximately 31.6% for fiscal year 2022 (OMB Payment Integrity Report), which has kept EITC returns as a persistent focus of the correspondence audit program under the W&I division.

Decision Boundaries

Selection by a DIF score or matching program does not automatically produce an audit. A classifier — an IRS employee trained in examination — reviews the flagged return to determine whether the potential tax adjustment justifies the examination resources. The Internal Revenue Manual instructs classifiers to consider both the dollar amount at issue and the likelihood of a change (IRM 4.19.1).

Two structural distinctions govern what follows selection:

Correspondence audit vs. field audit. A correspondence audit (IRS correspondence audits) is conducted entirely by mail and targets a limited number of line items — typically one deduction category or one unreported income item. A field audit (IRS field audits) involves an in-person examination by a revenue agent and covers the return comprehensively. Field audits are reserved for returns with complex issues or large proposed deficiencies; the threshold is not published as a fixed dollar amount but is applied judgmentally by the classifier and group manager.

Statute of limitations as a hard boundary. Under IRC § 6501, the IRS generally has 3 years from the later of the return's due date or filing date to assess additional tax. For returns with substantial omissions of income exceeding 25% of gross income reported, that window extends to 6 years. No statute of limitations applies to fraudulent returns or to years in which no return was filed. The IRS statute of limitations framework effectively defines the outer boundary of which tax years remain eligible for examination at any given point in time.

Once a return is selected, the taxpayer receives either a letter requesting documentation (for correspondence audits) or a formal appointment notice (for office or field audits). At that stage, the IRS notices explained framework governs what rights attach and what response deadlines apply. Taxpayers who disagree with audit findings have access to the IRS appeals process before the proposed assessment becomes final.