This is the first question most small producers ask: do I have to send my product to a lab? The short answer is no. FDA doesn't require lab analysis. It does hold you responsible for the accuracy of what you print, though, however you got the numbers.
Does FDA require lab testing?
No. FDA's regulations don't say how you have to arrive at your nutrient values. Its guidance, including the Nutrition Labeling Manual: A Guide for Developing and Using Data Bases, describes recipe calculation from reliable databases as one acceptable route.
Calculation does have one built-in limit. Added sugars can't be measured by a lab at all, so FDA expects records for those whatever method you use. See our added sugars guide.
How FDA checks your numbers
When FDA checks a label, it analyzes a composite of samples from a lot and compares the results with your label, using the compliance rules in 21 CFR 101.9(g):
- Class I nutrients are nutrients added in fortified foods, such as added vitamin D or iron. The lab value must be at least 100% of the label value.
- Class II nutrients are naturally occurring nutrients you're declaring, such as vitamins, minerals, protein, fiber, and other carbohydrates. The lab value must be at least 80% of the label value.
- Calories, total sugars, added sugars (when added sugars are the only sugars in the food), total fat, saturated fat, trans fat, cholesterol, and sodium: the lab value can be no more than 20% above the label value.
In plain terms: don't overstate the good stuff, and don't understate the stuff people limit. A label that rounds calories up a little is safer than one that rounds down. FDA also allows "reasonable excesses" of the good nutrients and "reasonable deficiencies" of the limited ones, within current good manufacturing practice.
Database calculation: how it works
- Weigh every ingredient in a batch, in grams, including water.
- Look up each ingredient's nutrients in a reliable source, such as USDA FoodData Central or your supplier's specification sheet.
- Add them up for the batch.
- Adjust for moisture loss during baking or cooking by weighing the finished batch.
- Divide by the number of servings, at the RACC-based serving size.
- Apply FDA's rounding rules, and build the label.
Strengths: - Fast and cheap - Easy to redo when a recipe changes - Supports the records FDA expects for added sugars
Weaknesses: - Database values are averages - Moisture loss and frying oil uptake are hard to estimate - Unusual ingredients may not be listed
Lab analysis: how it works
You send a few units of the finished product to an accredited food lab. The lab runs standard methods, usually AOAC, for the nutrients on the label, and reports per 100 g. You, or the lab, convert that to per-serving values and apply FDA rounding.
Strengths: - Measures your actual product, including cooking effects - Strong defense if a value is challenged - Handles complex or fried products well
Weaknesses: - Costs more, and turnaround can take weeks - A single sample may not represent batch-to-batch variation - Can't measure added sugars - Must be repeated when the recipe changes
Pricing varies widely by lab and panel. Ask two or three accredited labs for a quote on a full FDA Nutrition Facts panel.
Side-by-side comparison
| Database calculation | Lab analysis | |
|---|---|---|
| FDA accepts it | Yes, if accurate | Yes |
| Speed | Minutes to hours | Usually days to weeks |
| Cost | Low | Higher, varies by lab |
| Accuracy for simple recipes | Usually good | Very good |
| Accuracy for fried, dehydrated or fermented foods | Weaker | Better |
| Added sugars | From records (required) | Can't measure |
| Easy to update | Yes | Requires retesting |
When a lab test is worth it
- Fried foods, where oil uptake is unpredictable
- Dehydrated, baked-down, or reduced products, such as jerky, sauces, and dried fruit, where water loss changes everything
- Fermented products
- Products making claims, such as "high protein" or "low fat", where you sit near a legal threshold
- Wholesale to large retailers, which sometimes require lab data
- High-volume products, where the cost of an error is high
A sensible hybrid approach
Many small brands calculate first, then lab-test their best-sellers or anything making a claim. If the lab result and the calculation agree within FDA's tolerances, you've validated your method for similar recipes. If they don't, you've caught the problem before a regulator or customer does.
Keep a file for every product: - Batch weights - Ingredient sources - Database IDs or supplier specs - Any lab reports
FDA can ask for records for some nutrients, and good records make relabeling easy.
Where a label maker fits
Getting the numbers is one job. Laying them out in an FDA-compliant format is a separate one. Our builder handles the second part. You enter your per-serving values, from a lab report or your own calculation, and it produces the label in any of 13 FDA formats, plus Supplement Facts and the Canadian table. It applies the correct type hierarchy, rules, and footnote. Preview for free, and download a print-ready PDF for $4 when you're happy with it.
Getting the most from a database calculation
- Weigh, don't measure. Cup measurements of flour can vary a lot. Grams don't.
- Pick the most specific database entry: "wheat flour, white, all-purpose, enriched, bleached", not just "flour".
- Use supplier specs for branded ingredients like chocolate chips, protein powder, or sprinkles. They often differ from generic entries.
- Measure your yield. Weigh the batch before and after baking, and attribute the water loss correctly. Most of the weight lost is water, so nutrients concentrate.
- Version your recipes. Save each recipe with a date. When the label changes, you'll know why.
- Sanity-check against similar products. If your cookie comes out at half the calories of comparable cookies, recheck before you print.
Reading a lab report
Labs usually report per 100 g. To convert, multiply by your serving size in grams and divide by 100. Say a lab reports total fat at 18 g per 100 g, and your serving is 30 g. That's 18 × 30 ÷ 100 = 5.4 g per serving. Fat at 5 g or more rounds to the nearest gram, so the label shows 5 g. Do the same for every nutrient, then calculate %DV.
Frequently asked questions
Does the FDA require lab testing for nutrition labels?
No. FDA accepts values calculated from reliable nutrient databases, but you're responsible for their accuracy under the compliance rules in 21 CFR 101.9(g).
How accurate does my nutrition label have to be?
Naturally occurring vitamins, minerals, protein, and fiber must test at 80% or more of the label value. Added nutrients must test at 100% or more. Calories, sugars, fat, saturated fat, cholesterol, and sodium can't exceed the label value by more than 20%.
Can a lab measure added sugars?
No. Labs measure total sugars. Added sugars come from your recipe and supplier records.
Which is better for a home baker, software or a lab?
For simple baked goods, database calculation is usually enough. Consider a lab test for fried or dehydrated products, or any product making a nutrient claim.



