Probably not by much. In the most careful trial so far, adults who served themselves lunch on a small plate ate about the same as those given a large plate: 625 vs 644 kcal, a difference too small to tell from chance3. An earlier review of nine experiments found no consistent effect either1.
The idea is appealing: food looks like more on a small plate, so you feel satisfied with less. Here is what the tests actually show, and why portion size is the better-supported lever.
What was studied
The key trial ran in Cambridge, UK, in 20183. It had three features that older plate studies often lacked:
- Randomised. Each person was assigned a plate by chance, so the two groups started out alike.
- Preregistered. The researchers published their plan before collecting data, so they could not change the test after seeing results.
- Blinded. Participants did not know plate size was the point; they were told the study was about time of day and mental processes. An outside statistician who did not know the groups ran the analysis.
The 134 volunteers were adults aged 18 to 60 from the general public, about half men and half women3. Each ate in a room set up like a lounge, with a sofa, bookshelves and a nature programme on TV. They served themselves a cheese and tomato pasta bake from a heated trolley, as much or as little as they liked, over up to 30 minutes. The food was weighed before and after.
| What | Detail |
|---|---|
| People | 134 adults, aged 18–60 |
| Plates | plain white, 23 cm or 29 cm (about 59% more surface) |
| Meal | self-served pasta bake, one lunch |
| Measured | calories eaten, amount served, food left over |
| Design | randomised, preregistered, blinded |
In plain terms: a fair test of one question, with only the plate changed.
What the studies found
The careful trial: no clear difference. People with large plates ate 644 kcal on average. People with small plates ate 625 kcal3. The difference, 19 kcal, is about 3%. Statistically there was "no clear evidence of a difference".
Calories eaten at one self-served lunch, by plate size · kcal
The amount people served themselves was also similar: about 443 g on the large plates and 423 g on the small ones3. Meal length, bites and eating speed did not differ. The one difference: slightly more food was left on large plates (13.5 g vs 4.9 g).
The earlier review: no consistent effect. A 2014 meta-analysis (a study that pools many experiments) found nine experiments where people served themselves using bigger or smaller dishware1. Most found no significant difference. Pooled, there was a marginal tendency for bigger dishware to go with more eating, but it was small and the studies disagreed a lot. The authors concluded that smaller-plate advice "may be premature".
Other reviews found bigger effects. The trial's authors summarise them3. A Cochrane review estimated a small-to-medium effect of larger tableware. A 2016 meta-analysis estimated a large effect when food was self-served, and almost none when a fixed portion was put on different plates. The trial authors note that most of these older studies were judged low quality, and that pooling them "may have considerably overestimated effects".
Effect of larger sizes on food eaten, standardised · SMD
In plain terms: the best-designed test found little, and the bigger numbers come from weaker studies.
What this does and does not mean
"No clear difference" is not "zero". The trial's 95% confidence interval, the range the true effect most likely sits in, ran from 77 kcal less to 115 kcal more on large plates3. So a small effect either way is not ruled out. The authors add that even a 3% difference, if it held up in daily life, could matter across a whole population. They called the evidence "insufficiently conclusive to either endorse or refute" smaller-plate advice.
One meal, one food, one setting. The trial measured a single lunch of pasta bake. It says nothing about weeks of eating, other foods, or eating with family.
Smaller dishes can shift eating later. A 2023 lab study served 61 men a set breakfast in small, medium or large jars4. The small-jar breakfast left people feeling fuller and cut intake right after breakfast, but it was followed by more eating at the next meal, and self-recorded intake for the whole day came out higher than with the medium jar. That was one small study with food diaries, but it is a reminder that one meal is not a day.
Portion size is the stronger lever. A 2015 Cochrane review pooled 58 studies with 6,603 participants2. Larger portions, packages or tableware increased food eaten, with a small-to-moderate effect (standardised mean difference 0.38; in adults 0.46). Of its 72 included studies, about half changed portion size and about a fifth changed tableware. The plate-trial authors also note that cutting portion sizes "has a more robust evidence base"3.
In plain terms: what goes on the plate matters more than the plate.
What this means in practice
These nutrition studies describe settings, not rules. The size of the plate, on its own, changed little in the careful test. The amount served changed intake across many studies.
If you track food, the portion is also what you log. Weighing it is the most direct way to see it: see how accurately you need to track calories and whether to weigh food cooked or raw. For how a daily target is set in the first place, see how daily calorie targets are set.
The studies at a glance
| Study | Who was studied | Key result | Year |
|---|---|---|---|
| Robinson et al., *Obes Rev*1 | 9 experiments with self-served dishware (abstract) | Most found no significant difference; pooled effect small and inconsistent | 2014 |
| Hollands et al., *Cochrane*2 | 58 studies, 6,603 participants; portion, package or tableware size (abstract) | Larger sizes increased food eaten (SMD 0.38; adults 0.46), mostly from portion studies | 2015 |
| Kosīte et al., *Int J Behav Nutr Phys Act*3 | 134 adults aged 18–60, UK; randomised, preregistered | 625 vs 644 kcal eaten, small vs large plate; no clear difference | 2019 |
| Abeywickrema & Peng, *Foods*4 | 61 men aged 20–40; set breakfast in small, medium or large jars (abstract) | Small jars followed by more eating later; higher total daily intake | 2023 |
Limits worth knowing
- One meal. The best trial measured a single self-served lunch of one food3.
- Lab setting. A lounge built to look like a home is still a lab. Field trials in canteens and homes are needed, as the authors say.
- Adults 18 to 60 in the UK. Children and other settings were not tested.
- Wide confidence interval. A small effect in either direction cannot be excluded3.
- Abstract only for three papers. We read the 2019 trial in full; the 2014, 2015 and 2023 papers as abstracts.
- Older studies were weak. Most studies in earlier reviews were rated at high or unclear risk of bias2.
Latest research
The newest study we found is the 2023 jar study above4. It points the other way from the popular tip: a smaller dish was followed by more eating later in the day. It is small, in men only, and relies on food diaries after the lab meal.
We found no newer trial as rigorous as the 2019 plate study. It remains, in its authors' words, "the most robust study to date" on plate size and intake3.
The short version
- In the most careful trial, small and large plates led to about the same lunch intake: 19 kcal apart.
- A review of nine experiments found no consistent effect of dishware size.
- Portion size has much stronger evidence than plate size.
- A small plate effect, in either direction, is not ruled out.
Educational overview only — not medical advice.
