Inside the IRS Audit Playbook: How Revenue Agents Think, Investigate, and Decide

If you understand what an IRS Revenue Agent does on paper, you understand half of an examination. The other half — the half that determines outcomes — is how they think. The mental model an agent brings to a case shapes which issues get developed, which positions get pushed, which compromises get accepted, and ultimately whether your audit closes for $0, for the full proposed adjustment, or somewhere in between.

This is the IRS audit playbook from inside. Not the procedural manual published in the Internal Revenue Manual — that document is publicly available — but the working mental framework that experienced Revenue Agents actually use as they prioritize cases, identify issues, and make the dozens of small decisions that aggregate into an examination’s outcome.

This article is informed by Kugelman Law attorney Otto Bosch, who served as a Revenue Agent in the IRS Global High Wealth Group within the Large Business and International (LB&I) Division before joining the firm in February 2026. For an introduction to what Revenue Agents formally do and how examinations are structured, see our companion piece on what an IRS Revenue Agent does. The article below picks up where that one leaves off — focused not on the structure of an audit but on the thinking behind it.

How Returns Get on the Radar in the First Place

Most taxpayers imagine return selection as a uniform process. In practice, it is a triage. The IRS receives more than 150 million individual returns each year, and the examination function can pursue only a small fraction of them. Every return that reaches a Revenue Agent’s desk has survived multiple rounds of selection — which means by the time the examination opens, someone in the IRS has already decided this return is worth investing real resources in.

That triage happens through several mechanisms — DIF scoring, related-return pickups, information matching, project initiatives, and others — which we covered in detail in our prior article. What matters for understanding the playbook is the agent’s mindset when a case is assigned: the agent assumes there is something to find. The selection process is statistical, not certain — but it is good enough that experienced agents do not approach examinations as fishing expeditions. They approach them as recovery operations: the system has flagged something, and the agent’s job is to figure out what.

This default assumption matters defensively. Many taxpayer responses during an audit are calibrated to “look cooperative” or “explain things,” on the assumption that the agent is starting from neutral. The agent is not starting from neutral. The agent is starting from “the system thinks something is here.” Responses calibrated to that posture are different from responses calibrated to a neutral counterparty.

How a Revenue Agent Builds a Case from Suspicion to Adjustment

An audit, viewed from the agent’s seat, is not a single inquiry. It is a layered case-building exercise. Each layer corresponds to a different mental task.

Stage 1: Pre-Contact Intuition

Before the agent ever issues a notice, they have read the return, the notes from selection, prior-year filings, and any third-party data already in IRS systems. They have formed a working hypothesis about what the case is — and a working list of issues they expect to develop. Experienced agents are usually right about the rough shape of the case before the first IDR ever leaves their desk.

What the taxpayer sees as the “first contact” is, from the agent’s perspective, the third or fourth phase of the case. Defense strategies that treat the opening conference as the start of the audit are already a step behind.

Stage 2: Issue Identification Through Documents

The first Information Document Request (IDR) is the agent’s tool for confirming or refuting the pre-contact hypothesis. They are not asking for documents because they want to read receipts. They are asking because they want to see whether reality matches their hypothesis — and where reality does not match, they want to see where the gaps are.

Experienced agents read taxpayer responses for three signals: what was produced, what was conspicuously absent, and what the production reveals about how the taxpayer keeps records. A neat, well-organized response signals a sophisticated taxpayer (and probably a careful preparer). A messy, partial, or contradictory response signals issues that are likely to multiply as the audit goes deeper. Both responses tell the agent how aggressively to invest in the case.

Stage 3: Position Development

Once issues are identified, the agent shifts from finding things to building something. A “position” is the IRS’s articulated theory for why a particular adjustment should be made — and the case file the agent builds to support that position is what survives into Appeals, into Tax Court, and into any settlement discussion.

This is where mental discipline starts to differentiate experienced agents from inexperienced ones. Strong positions are built on documents, third-party records, and clean factual narratives. Weak positions rely on inference, taxpayer statements, or agent-developed math that the taxpayer can re-do. A good defense team can usually tell within the first few exchanges which kind of position the agent is building.

Stage 4: Workpaper Construction and Supervisory Sign-Off

Workpapers are not the agent’s notes. They are the IRS’s case file — the formal record that managers, IRS Counsel, Appeals officers, and (if it gets that far) the Tax Court will rely on. Every position the agent develops must eventually be expressed in workpapers that withstand internal review.

This creates a meaningful internal filter. Positions an agent personally believes in but cannot reduce to a clean workpaper get dropped. Positions a manager pushes back on get refined or abandoned. Positions IRS Counsel will not support get withdrawn. The defense team that understands this filter — that knows which positions are likely to survive review and which are not — can apply pressure exactly where it is most likely to produce results.

The Internal Pressures That Shape Every Audit Decision

A Revenue Agent does not have unlimited time, and the IRS does not have unlimited capacity. Every audit operates under three quiet but constant pressures that shape decisions taxpayers rarely see.

Cycle time. Agents have caseload expectations. An audit that drags is an audit that pulls the agent away from their other cases — and from their performance metrics. This is one of the reasons that responsive, well-organized taxpayer cooperation often produces better outcomes than passive resistance: the agent’s incentive is to close the case efficiently, and giving them a clean path to closure is sometimes worth more than fighting every issue.

Review risk. Every aggressive position the agent advances will be reviewed — by the manager, by IRS Counsel, sometimes by Appeals. An agent who advances positions that get overturned at review damages their internal credibility. This is why agents are often reluctant to push aggressive penalty positions, civil fraud allegations, or controversial legal theories unless the workpapers genuinely support them. Recognizing the threshold at which an agent will or will not commit to a position is one of the highest-leverage insights a defense team can have.

Specialty referrals. Complex examinations frequently involve specialists — international examiners, computer audit specialists, financial product specialists, valuation engineers. Bringing in a specialist takes time and case management. Agents weigh the value of escalation against its cost. A defense that signals a serious specialist would face credible counter-arguments may shift the case toward narrower issues that the agent can resolve without bringing in additional resources.

What Agents Look For That Taxpayers Don’t Recognize

Some of the most valuable inside-the-IRS knowledge is also the most counterintuitive. The signals below are things Revenue Agents are trained to read but that taxpayers and unprepared representatives often miss entirely.

• Lifestyle versus reported income. Significant gaps between what the return shows and what the taxpayer’s life suggests — homes, cars, travel, business interests visible on social media — are flags agents notice early. The IRS has access to public records and increasingly to other data streams that make these comparisons routine.
• Round numbers. Returns full of round numbers (exactly $5,000 in expenses, exactly $10,000 in donations) signal estimation rather than documentation. Agents notice this and adjust the audit accordingly.
• Inconsistencies across years. A line item that appeared in 2022 but vanished in 2023 — or a deduction that scaled non-linearly with income — invites questions. Agents do not always pursue these, but they note them, and they shape the case file.
• Related-party transactions without arms-length characteristics. Loans between entities with no documented terms, payments to family members for unspecified services, or rent to controlled entities at non-market rates draw immediate attention.
• Cash-intensive businesses with thin paper trails. Restaurants, salons, contractor businesses, and other cash-heavy operations get scrutinized differently. Agents are trained to test reported gross receipts against industry norms and against bank deposits.
• Crypto and digital asset patterns. Returns showing digital asset activity without corresponding income items, or returns answering “no” to the digital asset question while exchange data shows otherwise, are flagged. Our article on IRS cryptocurrency audits explores this pattern in detail.
• Suspiciously timed amendments and late filings. Returns amended after the IRS opened an audit, or returns filed unusually late after notices, draw heightened attention. Agents note timing.

The pattern in all of these: agents are reading the return for signals about the taxpayer, not just about the numbers. Defense strategies that focus only on document production miss this dimension entirely.

How Agents Decide Whether to Push or Fold on an Issue

One of the most useful insights from inside the IRS is the recognition that agents do not push every issue to its limit. Many issues are noticed, considered, and quietly dropped — because the cost-benefit math does not work for the IRS.

The internal calculus on a given issue weighs:

• The dollar amount at stake
• The strength of the documentary support
• The likelihood the position survives Appeals or Tax Court
• The agent’s confidence in the legal theory
• The amount of additional development needed
• Whether the issue connects to other issues already being developed

Issues with strong documentary support, clear law, and meaningful dollars tend to be pushed. Issues with weak documentary support, ambiguous law, or trivial dollars tend to be dropped — even if the agent personally suspects the taxpayer’s position is wrong. The IRS does not pursue every theoretical adjustment. It pursues the ones that pencil out.

A sophisticated defense uses this. By making strong positions stronger and exposing weak positions early, the defense can shift the agent’s calculus on a case. Issues that were borderline tend to fall toward dropping. Issues that were marginal tend to settle on terms favorable to the taxpayer.

When the Audit Plays By Different Rules

Most of the playbook described above applies to standard examinations. There are categories of audits where the rules shift, and recognizing the shift is critical.

Eggshell audits — civil examinations with potential criminal implications — operate under an entirely different set of rules. The agent’s job is no longer to develop adjustments efficiently but to develop the record carefully, with an eye toward potential referral to IRS Criminal Investigation. Cooperation strategies that make sense in a routine audit can be catastrophic in an eggshell audit.

Global High Wealth and LB&I enterprise audits are also their own world. Cycle-time pressures are different, specialist resources are abundant, and the audit considers the entire web of related entities and transactions rather than the individual return. The mental model a Global High Wealth team brings to a case is integrated and patient in ways most taxpayers do not expect.

Project-driven examinations — audits opened as part of a focused enforcement initiative — also play differently. The agent has trained on the project’s target issue, has examined other taxpayers in the same project, and has internal guidance on what positions to develop. A defense that does not recognize the project’s contours will misread the agent’s posture entirely.

In each of these scenarios, defending without inside-the-IRS perspective is defending blind.

What This Means for Defense Strategy

The aggregate of everything above has a single practical implication for taxpayers under audit: the most consequential decisions in your audit are not the visible ones. They are the unseen ones — the agent’s pre-contact hypothesis, the position-building decisions in the workpaper file, the internal review pressures, the issue-by-issue cost-benefit calculations, and the specific signals the agent is reading from your responses that you do not realize you are sending.

This is what an IRS-insider perspective on the defense team actually changes. With Otto Bosch’s experience inside the IRS Global High Wealth Group and Alex Kugelman‘s nearly two decades of federal tax controversy litigation, Kugelman Law approaches every audit defense matter with a working understanding of the playbook on the other side of the table — and a credible litigation backstop if the case cannot be resolved administratively. We covered the full team capability in our article on why a former IRS revenue agent attorney changes audit defense.

Representative outcomes from the firm’s audit defense practice include a $365,000 tax debt reduced to a zero-dollar liability, a multi-year audit and non-filing matter resolved with minimal payment, and ten years of unfiled returns brought into compliance with a successful outcome. Results depend on specific facts. Past results do not guarantee future outcomes.

If you would like to discuss your IRS or FTB matter and how the firm’s combination of inside-the-IRS perspective and federal tax litigation experience can shape your defense, see our tax help resources or contact the firm directly.

Frequently Asked Questions

How do IRS auditors decide which issues to focus on?

Revenue Agents prioritize issues based on a combination of dollar magnitude, strength of documentary support, clarity of the legal theory, and the cost in agent time required to develop the position. Issues that are well-supported, technically clean, and material to the case tend to be pushed. Issues that are weak, ambiguous, or trivial tend to be quietly dropped — even when the agent personally suspects the taxpayer’s position is incorrect.

What red flags do IRS auditors look for?

Common signals include lifestyle inconsistent with reported income, returns full of round numbers, year-over-year inconsistencies in reported items, related-party transactions without arms-length characteristics, cash-intensive businesses with thin documentation, digital asset activity that does not align with reported income, and suspiciously timed amended or late-filed returns.

Can an IRS auditor decide to drop an issue mid-audit?

Yes. Issues that look promising in pre-contact analysis frequently get dropped during fieldwork as documents and explanations come in. Conversely, issues that were not initially identified can emerge from the development process. Audit scope is not fixed at the opening conference — it evolves as the case develops.

What does it mean when an IRS audit closes “no change”?

A no-change closing means the agent did not develop adjustments and the return is accepted as filed. This outcome is more common than many taxpayers assume. It occurs when the issues identified at selection do not survive document review, when the taxpayer’s documentation is strong, or when the cost-benefit math on the available positions does not justify pursuing them.

How do I know if my IRS audit is becoming an eggshell audit?

There are signals — agent questions that focus on knowledge, intent, and willfulness rather than documentation; involvement of specialized fraud or referral-related personnel; specific timing patterns in document requests; and sudden agent reluctance to discuss the case. Recognizing these signals reliably requires controversy experience. If you have any reason to suspect criminal exposure, attorney representation is essential and should be retained before any further communication with the IRS.

Speak With Kugelman Law

If you are facing an IRS audit, controversy, or complex federal tax matter — or if you suspect the IRS is preparing to open one — schedule a paid privileged consultation with Kugelman Law. Call (415) 968-1780 or visit our contact page. All consultations are fully protected by attorney-client privilege.