Food-label investigation · United States · Checked November 19, 2026
The zero on a cooking-spray label is a rounded number attached to a tiny serving. It is not permission to multiply that zero into an unlimited amount of oil. A food diary that keeps the zero but drops the serving has discarded the most important part of the label.
This is not evidence that a manufacturer has hidden a secret calorie count. It is a practical problem created when a display convention becomes an arithmetic input. We checked a specific manufacturer listing, then built a worksheet showing what its rounded declaration can and cannot establish.
The serving is the story
Conagra's US foodservice listing for PAM Original Cooking Spray, case of six 17 oz cans, declares a 0.25 g serving, 0 calories and 0 g total fat. That is the exact pack and source used here. We did not substitute a grocery-store can, a different oil blend or a retailer's transcription.
| Product identity | Declared serving | Declared calories | Declared total fat |
|---|---|---|---|
| PAM Original, US foodservice, 6 × 17 oz case | 0.25 g | 0 kcal | 0 g |
The listing identifies the serving in grams. We have not established a measured output in grams per second for this can. A spray duration from another package is not a calibration of this one. That distinction matters whenever someone turns a quick press of a nozzle into a supposedly exact food entry.
Why zero can still represent some energy
Under 21 CFR 101.9(c)(1), a serving with fewer than 5 calories may be expressed as zero on Nutrition Facts. The same regulation allows total fat below 0.5 g per serving to be expressed as zero. These are separate reporting thresholds, not statements that every molecule of fat or calorie has disappeared.
For a mathematical illustration, imagine a serving containing 2 calories before rounding. Ten such servings contain 20 calories before rounding. Adding ten printed zeros gives zero. The disagreement arises because rounding first and multiplying later need not produce the same answer as multiplying the underlying quantity first.
We are not asserting that this PAM serving contains exactly 2 calories. That value is deliberately hypothetical. The manufacturer declaration alone does not supply an unrounded calorie value.
What you can calculate without inventing a measurement
| Number of declared servings | Mass at 0.25 g each | Printed zero multiplied | Bound from the calorie-rounding rule alone |
|---|---|---|---|
| 1 | 0.25 g | 0 | Less than 5 kcal |
| 4 | 1 g | 0 | Less than 20 kcal |
| 20 | 5 g | 0 | Less than 100 kcal |
| 40 | 10 g | 0 | Less than 200 kcal |
These deliberately loose bounds are not calorie estimates for those portions. They use only the rounding threshold and an equal-serving assumption. Composition could make a much tighter estimate possible; the printed zero alone cannot. The table is a demonstration of lost information, not a claim that 10 g of this spray contains nearly 200 calories.
It would be just as misleading to publish those upper bounds as a sensational warning about what people “really” consume. A bound is not a measured value. We have neither assayed the spray nor measured how much remains on food after cooking.
Oil mass is a better model than nozzle time, with limits
A separate hypothetical model uses 5 g of pure oil treated as fat at the general factor of 9 kcal/g. That gives approximately 45 kcal; 10 g gives approximately 90 kcal. The calculation assumes a known mass of oil and that all of it is included in the portion being estimated. It is not an ingredient analysis of PAM and does not account for material left in the pan.
That is why weighing a can before and after spraying is not automatically a measurement of oil eaten. Container mass loss, material reaching the pan and material consumed are different quantities. A scale also needs enough resolution for the amount in question. Writing more decimal places in a diary does not repair a measurement the scale could not resolve.
For ordinary logging, use an entry that matches the actual product and the best available portion information. If you know the mass of plain oil added to a recipe, record that as an estimate with its assumptions. If you know only an approximate spray duration, keep the uncertainty visible. Do not claim a precisely weighed amount you did not measure.
Three questions a food app should preserve
- What does the zero refer to? Keep the declared serving mass beside the number.
- What portion is being logged? A package serving, a timed spray and a weighed quantity are different kinds of evidence.
- Where did the unrounded estimate come from? Manufacturer data, a generic oil model and a laboratory assay should not carry the same confidence label.
The larger lesson extends beyond cooking spray: a display value is not always a suitable calculation value. Our calorie-versus-macro investigation addresses a different problem, reconstructing energy from rounded nutrient fields. Here the error is multiplying a rounded zero while forgetting its very small serving.
Method, downloads and limits
CSV declarations and worked examples · JSON with formulas and sources. We manually reviewed the manufacturer listing and US rule on November 19, 2026. This is one documented product example, not a market survey, a pouring experiment or a chemical test. Package labels should take precedence when they differ from the online listing. No price comparison or health outcome is implied.
Publisher disclosure: we make Vora. Vora combines proactive AI with nutrition, workouts and wearable context. If a lower-friction diary would help, see supported text-based logging workflows. A useful assistant should retain what the portion means, including when it is an estimate. The worksheet is available independently of the app. Send source-specific corrections to support@askvora.com.