CDTFA Audit Sampling: How to Review a Test Period and Dispute Unsupported Results
ARC · California sales & use tax consulting
A small group of transactions can have a large effect on a CDTFA sales tax assessment. When an audit uses sampling, the important questions go beyond whether a particular invoice is taxable. You also need to understand which transactions were selected, what they represent, and how the results are applied to the rest of the audit.
Yes. Ask for the sampling plan, understand the population being tested, and identify specific concerns supported by business records. A useful objection explains which transaction, period, assumption, or calculation needs review—and why.
Sampling itself is not evidence of an improper audit. The goal is an accurate result. For help evaluating the method and findings, explore ARC’s California sales tax audit defense and representation services.
What is CDTFA audit sampling?
Sampling examines selected transactions to draw conclusions about a larger group. CDTFA’s Publication 76, Audits, distinguishes an actual-basis review of every transaction from a sample-basis review. It explains that the choice depends on the business and records, and that more than one approach may be used in an audit.
For a business owner, the practical distinction is simple: a finding about a selected transaction may affect more than that transaction. You therefore need to follow the calculation from the source record through the proposed adjustment.
Key terms to clarify with the auditor
- Population: the defined group the sample is intended to represent.
- Sample: the selected transactions or other units examined.
- Test period: a period used in the testing approach; not every sampling method selects one continuous block of time.
- Projection: application of the test findings to a broader group.
- Actual review: examination of each transaction within the defined scope.
Ask how these terms are being used in your particular audit. A statement that “we are using a sample” does not identify the method, population, or projection.
Why the projection deserves careful attention
A mistake in classifying one selected item can affect the test result. An error in the population or projection can affect the broader calculation. Review both the individual findings and the method used to extend them.
Suppose a hypothetical test identifies $5,000 of additional taxable sales in a $100,000 sales base. A simple ratio is 5%. Applying that ratio to a $2,000,000 comparable sales base produces $100,000 of additional taxable sales.
That $100,000 is not the tax bill. It is an illustrative taxable-sales adjustment before applying tax rates, credits, or other relevant calculations. Real audits can use different methods and definitions.
The example shows why it matters to ask whether the selected items, ratio, and application base are appropriate. It does not establish that a sample is wrong or that changing one item necessarily produces a proportional reduction in an actual assessment.
Seven questions to ask about a CDTFA sampling plan
CDTFA’s Audit Manual, Chapter 4, sections 0405.20 and 0405.23, discusses test periods, taxpayer participation in developing the plan, and Form CDTFA-472, Audit Sampling Plan. The manual describes documenting both the sampling method and projection.
Use the following questions as a preparation agenda. They are not automatic grounds for rejecting a sample.
| Question | What to clarify |
|---|---|
| What issue is being tested? | Sales, purchases, a deduction, or another defined reporting question. |
| What is the population? | Accounts, locations, dates, transaction types, and exclusions. |
| How were items selected? | The selection method and the records used. |
| Why does this test fit the business? | Relevant operational facts and changes during the audit period. |
| How are findings classified? | The facts and rules supporting each proposed error. |
| How are results projected? | The formula, denominator, application base, and covered periods. |
| How will corrections be handled? | The process for addressing new documents or calculation issues. |
Keep the explanation with the plan and schedules. If a term or calculation remains unclear, ask for clarification before treating a summary number as the complete answer.
What if the test period does not reflect normal business activity?
An unusual period is a reason to investigate, not an automatic reason to discard the test. Identify the change, its dates, and how it affects the specific issue being measured. The strongest explanation connects operating facts to the proposed calculation.
Business changes worth investigating
- A location opened, closed, or operated for only part of a period.
- A software migration changed how reports were generated.
- A temporary promotion changed the mix of transactions.
- A major customer or product line entered or left the business.
- A one-time purchase appears alongside recurring expenses.
- Seasonal activity differs across the periods being compared.
These are review prompts, not claims that any event makes sampling invalid. Assemble dated records and explain the connection. A statement that “this was a bad month” is less useful than a documented explanation of which transactions changed and why that matters.
Do not choose a replacement period only because it produces less tax
If you propose another approach, explain why it better addresses the issue. Present the relevant facts consistently, including facts that do not favor your position. A credible review aims to establish the correct treatment rather than select whichever result looks smallest.
Review the transaction findings as well as the sample design
A sound sampling approach does not eliminate the need to verify each proposed error. Match the selected item to its original invoice, description, tax treatment, and supporting documents. Record whether your concern involves a factual detail, tax classification, duplicate item, timing, or computation.
For specialized transactions, explore ARC’s manufacturing sales tax exemption services, food processing exemption services, and winery sales and use tax consulting. The applicable treatment depends on the particular facts and law.
Preserve the source documentation. If a selected record is unavailable, identify the gap and discuss appropriate support; do not fabricate an original document or assume a different invoice automatically proves the missing transaction’s treatment.
How to build a documented objection to sampling results
Use a concise issue schedule rather than a general letter saying the assessment is excessive. This is a practical organization method, not a prescribed filing form.
- Identify the reference. Name the schedule, item, or calculation you are questioning.
- State the issue. Explain the exact fact or assumption that needs review.
- Attach the evidence. Connect each document to the disputed point.
- Explain the proposed correction. Show what should change and why.
- Track the response. Record whether the issue was accepted, rejected, or remains open.
Instead of writing “the sample does not reflect our business,” a company identifies a location included in the projection base before it opened. It supplies the opening date and records supporting the location’s actual activity. The reviewer can then evaluate that specific population question.
Regulation 1698.5, Audit Procedures, addresses audit conferences, progress communication, findings, and information requests. Use the audit process to raise questions promptly and keep a record of the documents you provide.
Do not confuse an audit discussion with a formal appeal
If a Notice of Determination arrives while sampling questions remain unresolved, review its separate petition instructions immediately. For covered sales and use tax determinations, Revenue and Taxation Code section 6561 sets a 30-day petition period after service. A request to revisit a calculation should not be treated as a substitute for a required timely filing.
Put Decades of California Sales Tax Experience Behind Your Audit Review
Before you accept a projected assessment, understand what the sample actually supports. ARC’s California sales tax audit defense and representation services help businesses evaluate audit questions and prepare documented responses.
Joan Armenta-Roberts brings experience from both sides of the audit process: 17 years with the California State Board of Equalization, including work as a tax auditor, followed by 18 years in public accounting at KPMG and PwC. That background informs ARC’s approach to examining findings, identifying disputed issues, and presenting the records and explanations that support your position. Meet Joan and the ARC team.
Your review should go beyond the total assessment. The selected transactions, business history, supporting documents, and projection each deserve attention. If the records also raise potential overpayment issues, discuss whether a separate sales tax refund or reverse audit review is appropriate.
Get an experienced review before you accept findings you do not understand. Contact ARC to discuss your CDTFA sampling questions, or call Joan at 916-947-5160 or Zach at 213-453-1816. Have your audit correspondence, sampling plan, and response dates available.
Frequently asked questions about CDTFA audit sampling
Can CDTFA use a sample instead of reviewing every transaction?
Yes. Publication 76 describes both sample-basis and actual-basis reviews. The appropriate approach depends on the business, records, and issues.
Does every sample use a single test month?
No. Ask how the specific selection works. A statistical selection and a continuous test period are not necessarily the same approach.
Can I ask how the sample was selected?
Yes. Request an explanation of the population, selection method, and projection, and review the sampling plan.
Does an unusual month automatically invalidate the test?
No. Explain the relevant difference and support it with records. Its effect depends on the issue and method being evaluated.
Can one transaction affect more than its own tax amount?
In a projected result, an item’s treatment can affect the broader calculation. Trace the actual method rather than assuming a particular numerical effect.
Should I agree because the calculation looks mathematical?
Understand the inputs, definitions, and supporting facts as well as the arithmetic. A correct multiplication does not by itself establish the underlying tax treatment.
What if a formal determination arrives during the discussion?
Review the notice’s response instructions immediately. Ongoing audit discussions should not be assumed to preserve formal petition rights.
When should I contact ARC?
Contact ARC when you need help understanding a proposed test, reviewing findings, or preparing a supported response. Identify any approaching deadline at the outset.
Get Clarity on the Method Behind Your Assessment
A projected number deserves a documented explanation. Talk with ARC about the sampling plan, transaction findings, and questions your business needs to resolve before its next response.
Joan: 916-947-5160
Zach: 213-453-1816