3 in 4 foods warned per 100 g, 6 in 10 per serving
Of 349 foods with a usable serving, 265 are warned per 100 g and 214 are warned when the same limits are put on one printed serving.
How to check a pack yourself
- Look at the nutrition table. The nutrition table has a per 100 g (or 100 ml) column and often a per serve column.
- Compare using per 100 g. To compare two packs, use the per 100 g column, because each pack picks its own serving.
- Check your serving size. Look at the printed serving size and compare it with how much you actually eat.
Quick answers
What did the food industry ask the Supreme Court?
The All India Food Processors' Association asked for greater consideration of the amount consumers actually eat in one serving, instead of a standard 100 g for solids and 100 ml for liquids. FSSAI's own minutes from 2021 show that the industry suggested a per serve basis. FSSAI kept 100 g/ml as the reference unit because it 'enables a rational comparison among products'. The 2020 labelling rules already ask for the per serve RDA line, which shows how much of a day's recommended amount one serving gives. In July 2024 FSSAI announced its Food Authority approved showing it in bold, bigger letters. A draft dated 17 February 2025 proposed that for added sugar, saturated fat and sodium. On 27 October 2026 it was still not final on FSSAI's own lists. It is a separate rule from the warning label in the court case.
Why does a smaller serving mean fewer warnings?
If a food has a smaller serving size printed on the pack, the amounts of sugar, fat and salt per serving are smaller. This means the food may not meet the limits for a warning under the per-serving test. 53 foods lose the warning, all of them solid foods, with a middle printed serving of 17 g. Biscuits go from 33 of 33 warned per 100 g to 22 of 33 per serving, so most biscuits are still warned.
Why do drinks barely change?
The middle printed drink serving is 200 ml and only 6 of 65 drink servings are under 100 ml, so one serving holds as much as 100 ml or more. 49 of 65 drinks are warned per 100 ml and 50 per serving.
Which column should I read at the shelf?
To compare two packs, use the per 100 g (or 100 ml) column. Each pack picks its own serving size. This column shows the same standard, so it is easier to compare.
Is the warning in force?
The warning is a proposal, not yet law. The Supreme Court has not made a decision yet. The plan is being considered, but it is not in effect.
Every measure
| Measure | Share of foods | Foods |
|---|---|---|
| 349 foods, per 100 g | 76% | 265 of 349 |
| 349 foods, per serving | 61% | 214 of 349 |
| Biscuits, per 100 g | 100% | 33 of 33 |
| Biscuits, per serving | 67% | 22 of 33 |
| Sauces, per 100 g | 71% | 30 of 42 |
| Sauces, per serving | 52% | 22 of 42 |
How I did it
The study uses the same 665 Indian packaged foods as study 3. I re-ran the test with the same rules and ingredient matchers. Added sugar is the label's declared added sugar if present, or total sugar if a sugar name is in the list. Added fat is total fat if an oil or fat is listed. Salt is salt if listed. Per serving is the per 100 g value times the printed serving divided by 100. The code checks the result two ways and they agree on every food. The lists and servings come from Open Food Facts, entered by volunteers, so they are not a random sample. The serving is what the pack prints, as typed in by a volunteer. Servings reading exactly 100 g or 100 ml are left out. The limits were set per 100 g and 100 ml; the per-serving test is a what-if. FSSAI's plan is a proposal, not yet law, and the Supreme Court order is reserved. Fat is an upper estimate. No brands are named.
Where this comes from
- FSSAI: minutes of the stakeholders' meeting on front-of-pack labelling, 30 June 2021 (letter of 29 October 2021)
- The Logical Indian, 10 September 2026: the food processors' association's Supreme Court filing
- The South First, 17 August 2026: FSSAI's 2021 records
- PIB, 6 July 2024: FSSAI approves bold, bigger nutrition labelling (44th Food Authority meeting)
- FSSAI: draft Labelling and Display Amendment Regulations, 2025 (17 February 2025)
- LiveLaw: FSSAI's 28 August 2026 affidavit and the ICMR-NIN limits
- ThePrint: FSSAI's single-phase plan (23 September 2026 affidavit)
- Open Food Facts
What this can't tell you
- The lists and servings come from Open Food Facts, entered by volunteers, so they are not a random sample.
- The serving is what the pack prints, as typed in by a volunteer; servings reading exactly 100 g or 100 ml are left out.
- The limits were set per 100 g and 100 ml; the per-serving test is a what-if.
- FSSAI's plan is a proposal, not yet law, and the Supreme Court order is reserved.
- Fat is an upper estimate.
Data and code
- Every number on this page with the count behind it (JSON) and the same as a table (CSV)
- Code: https://github.com/devnarsinghani22/india-label-index (per_serving.py)
- Licence: the data comes from Open Food Facts under the Open Database License. You can reuse these numbers with credit to Open Food Facts and this page.
Part of the India Label Index 2026.
Narsinghani, D. (2026). Warning shares per 100 g vs per serving. India Label Index 2026. https://devnarsinghani22.github.io/india-label-index/per-serving/. Data from Open Food Facts (ODbL). DOI: 10.5281/zenodo.23245061.