What does a second review of state tax returns look for, and when is it worth doing?
Edvin Givargis Published 7 minute read
The short answer
A second review is a fresh examination of filed state returns, performed by someone who did not prepare them, looking for both overpayments that support refund claims and understatements that deserve correction before an auditor finds them. The systematic findings cluster in four areas: apportionment factor composition, state modifications to federal income, the composition of the filing group, and the tracking of attributes such as net operating losses and credits across years and entities. The review is most valuable at specific moments: before a refund statute of limitations closes on a profitable year, after a major transaction or law change, when an exposure has surfaced and the size of the understatement itself drives penalties, and when a company's returns have been prepared the same way, by the same process, for many years.
Why filed returns hold findings
State returns are usually prepared under compliance deadlines, by a process that starts from the prior year and from federal numbers, and that process is good at consistency and weak at reexamination. Positions set years ago roll forward because they are in the workpapers, not because anyone has recently tested them against current law. The states, meanwhile, change constantly: apportionment formulas move, nexus standards shift, conformity dates update, and case law reinterprets statutes that the workpapers still apply as originally read. The gap between how a return was prepared and how the law now reads is where a second review lives, and the gap widens every year the process runs unexamined. The same gap runs in both directions, which is why a review that only hunts refunds, or only hunts exposure, sees half the picture. The findings that matter most are often the ones that net against each other within a year or across a group.
The four places to look
Apportionment comes first because it moves the most money. Factor composition questions, what belongs in the sales factor and at what amount, where receipts from services and intangibles are sourced, whether thrown back sales still meet the throwback test under current nexus law, whether excluded items such as treasury receipts or occasional sales were properly kept out, routinely change state taxable income by more than any other category. Sourcing rules for services and intangibles have been rewritten in most states within the past decade, and returns that carried forward older sourcing conclusions are the single richest vein a reviewer works.
Modifications come second. Every state builds its base by adjusting federal income, and the adjustments are where federal changes collide with state conformity: bonus depreciation decoupling, interest limitation differences, dividend received treatment, related party addbacks and their exceptions, and the treatment of items the state never conformed to at all. Modification errors compound because they often affect an attribute, so a mistake in one year misstates every later year that draws on the carryforward.
Group composition comes third. Which entities file together, and on what basis, is a question companies answer once and then stop asking. Acquisitions bring in entities whose unitary status was never analyzed, dispositions leave members in the group after the facts that justified inclusion are gone, and elections that made sense at the time, worldwide or water's-edge, consolidated or separate, sit unrevisited while the footprint that drove them changes. Composition errors are the most expensive kind when they surface on audit, because they restate the entire computation rather than one line, and they are also where legitimate offsets hide: an entity that should have been in the group can bring losses and factors in with it.
Attributes come fourth. Net operating losses, credits, and deferred items each carry their own limitation periods, sharing rules, and adjustment mechanics by state, and schedules that track them are the most error-prone workpapers in the state file. A reviewer traces each material attribute from the year it arose to the year it is used, in each state, under each state's own rules rather than the federal schedule, because states diverge on carryforward periods, suspension years, ownership change limits, and how much of a loss survives a reorganization.
The moments that justify the work
Timing decides whether findings are worth anything. A refund claim needs an open statute of limitations, typically three to four years from filing or payment depending on the state, so the review of a profitable year is worth the most just before that window closes and worth nothing after. A company that reviews on a cycle tied to its largest states' limitation periods captures what a one-time review would forfeit.
Transactions and law changes create the second moment. A major acquisition, disposition, or restructuring changes group composition, factor geography, and attribute usage all at once, and the first return filed after the transaction is where those changes are most likely to be handled by rolling forward a process built for the old structure. Similarly, when a state rewrites its apportionment or conformity rules, the transition years deserve a look on their own.
The third moment is defensive, and it is the least understood: when an exposure has already surfaced. Once a company knows a filed return understated tax, the instinct is to quantify the error and reserve or amend for it. The better instinct is to review the entire year first, because the exposure is measured on the net understatement, and offsets found in the same year, an overstated factor numerator, a missed modification, an unclaimed credit, reduce the exposure dollar for dollar. Where a state imposes penalties that switch on at a fixed understatement threshold, and some strict liability penalties cannot be waived or negotiated at all once the threshold is crossed, the review can matter beyond the tax: bringing the corrected liability down can bring the understatement below the threshold and eliminate a penalty that no amount of reasonable cause argument could touch. A company that concedes the exposure at its first-computed size, without testing the rest of the year, may be paying a penalty the facts never required.
What a second review is not
A second review is not a promotion of aggressive positions into filed returns, and the distinction is worth drawing precisely. The review tests filed positions against the law as it stands: where the return claimed less than the law allows, the finding supports a refund claim the state's own procedures exist to handle, and where the return claimed more, the finding supports a correction the company controls the timing and presentation of. Both halves rest on positions the reviewer would defend in examination. A review run to a refund quota produces claims that fail on audit and sour the state relationship; a review run honestly produces a net position the company can stand on, whichever direction it points. It is also not a substitute for the compliance process. The findings feed back into the process that prepares next year's returns, and a review whose corrections do not change the following year's workpapers has fixed the past while re-purchasing the same findings for the future.
Practice notes
The review file has a shape. First, sequence by statute of limitations: list the material states, the open years in each, and the dates the refund windows close, and let that schedule set the order of work, because a finding in a closing year is worth more than a larger finding in a year with time. Second, start from the largest apportionment states and work the factor before anything else: request the sourcing workpapers, not just the returns, since the errors live in how the numbers were built. Third, run the group composition question fresh rather than confirming the existing answer: list every legal entity, its activity, and its relationship to the filing group, and require a current reason for each inclusion and exclusion. Fourth, when the review is defensive, compute the net understatement under the applicable penalty thresholds before and after offsets, and document the offsets with refund-claim rigor even though they are being used to reduce an exposure, because they will be tested exactly that hard when the state examines the amended year.
This article states the law as of September 14, 2026
Statutes, rates, thresholds, and agency practice change, and a different set of facts can change the answer. Before acting on anything discussed above, contact G&G State Tax Group to confirm what has changed since this was written and how the rules apply to a specific situation.
G&G State Tax Group, LLC is a state and local tax advisory firm. The firm provides state and local tax consulting and representation in state and local tax controversies. G&G does not prepare or file tax returns, perform attest services, or provide bookkeeping, and is not a CPA firm.