ABSTRACT In two studies, people estimated the percent of vegans, vegetarians, those following a reduced meat diet and omnivores for specific populations. In a study of Polish adults ( n = 1038), participants provided estimates of the diets of Poles and of residents of the US. In a US student sample ( n = 2538), participants provided estimates of the diets of their fellow students and of the US population. With a few exceptions, participants overestimated the sizes of dietary minorities (vegans, vegetarians, and reducetarians) and underestimated size of the dietary majority (omnivores). Moreover, these under‐ and overestimates did not vary systematically as a function of participant diet or vegetarian threat. Although women's estimates of the size of dietary minorities were larger than men's and women's estimates of the size of the omnivorous majority were smaller than men's, both men and women overestimated the percent of dietary minorities and underestimated the percent of omnivores. These misestimates reflect the operation of a general cognitive bias called ‘Uncertainty‐Based Rescaling’, that has been found to explain why people consistently underestimate the sizes of majorities and overestimate the sizes of minorities without the need to infer the operation of biases due to similarity, threat and so forth.