How Florida’s Sales Tax Audit Sampling Works for Restaurants?

In a Florida restaurant sales tax audit, the Department rarely reviews every transaction. Instead it examines a sample, calculates an error rate, and projects that rate across your entire audit period, usually three years. Florida law under section 212.12 allows this when records are adequate but too voluminous to review in full. The danger for restaurants is seasonality: if the sample lands on a holiday week or an unusual month, an unrepresentative result gets multiplied across years. A small sample error can become a large assessment, which is why the sample period and the agreement you sign matter as much as the underlying numbers.

Restaurants generate huge transaction volumes, so Florida auditors almost never count every sale. They sample. That single decision, how the sample is chosen and whether you agree to it, often shapes the assessment more than your actual compliance does.

Why Florida Uses Sampling?

Under section 212.12, Florida Statutes, when a business has adequate but voluminous records, the Department may examine a sample and project the findings over the full audit period. The goal is to estimate the proportion of taxable sales to total sales, or taxable purchases to total purchases.

Before sampling, the Department is supposed to make a good-faith effort to agree with you on the method. If no agreement is reached, you are entitled to a review by the executive director. That right exists for a reason: the method drives the number.

The Two Kinds of Sampling

Florida uses two broad approaches, and which one applies depends largely on whether you have usable electronic records.

Attribute	Stratified Statistical Sampling	Block / Judgment (Non-Statistical)
When used	When adequate electronic records exist	When electronic data is limited
How the sample is chosen	Data split into strata, then randomly sampled	Selected blocks of time, often a few months
Precision	Calculated, often to a 95 percent confidence level	No statistical precision calculated
Seasonal risk	Lower, if strata are sound	High, a peak or slow month skews everything
Your leverage to challenge	Question the strata and method	Argue the periods are unrepresentative
Measurement (how it becomes your bill)	Sample error rate is projected across the full period with a precision range	Sample error rate is applied straight across the full period, with no precision check

The Block Sample Problem for Restaurants

A common block sample is three months, one from each year of the audit period. The error rate found in those months is then projected across all three years.

For a seasonal business, that is dangerous. A South Florida restaurant booming in season looks very different in August. If the sample captures an unusual stretch, a holiday rush, a slow month, a period with a POS glitch, the projection magnifies that anomaly across the entire audit. One bad sample month can become a three-year assessment many times larger than any real shortfall.

The Sampling Agreement Decision

Early in the audit, the auditor will usually ask you to sign a form approving the sampling methodology. The state can sample whether or not you sign, but the signature is not a formality.

Approving the method can limit your ability to later argue that the chosen periods were not representative of your business. That is a meaningful right to give up, and it is a decision better made with someone who can first judge whether the proposed sample actually reflects your operation.

Evaluating the proposed sample, the periods selected, and the error-rate math before you agree to anything is exactly the work that happens inside a Business CPA Tax Resolution Case Analysis, where the methodology is tested against your real seasonality before it hardens into an assessment.

Where Restaurants Can Push Back?

A sample-based assessment is not the final word. The leverage points generally include:

• Representativeness: showing the sample months do not reflect your normal operations.
• Method: questioning how strata were built or why a block period was chosen.
• Error-rate validity: challenging how individual sample errors were classified.
• The DR-1215 stage: preliminary findings arrive as a Notice of Intent to Make Audit Changes, a key window to respond before the assessment is final.

Knowing which of these applies to your numbers, and documenting it correctly, is where a projection gets reduced. It is also where most owners, working alone and under time pressure, fall short.

Common Mistakes Restaurants Make

• Signing the sampling agreement without checking whether the periods are representative.
• Letting the auditor pick sample months that ignore seasonality.
• Accepting the projected error rate without auditing the individual sample items.
• Missing the DR-1215 window to challenge preliminary findings.
• Assuming a sample result is too technical to dispute.

Questions Owners Actually Ask

Should I sign the sampling form? Can I pick which months they test? The auditor used my busiest month, is that allowed? Those questions go to the heart of the assessment, because in a sampled audit the method is the case.

Start with what a Florida DR-840 audit notice means and what happens next, and see how the markup method often rides on the same sampled data. When the projection is the problem, a CPA review of the FDOR calculation is the next step.

Frequently Asked Questions