The manuscript for this project was accepted at Cognition: https://doi.org/10.1016/j.cognition.2022.105149 Abstract: Children struggle with the quantifier "most". Often, this difficulty is attributed to an inability to interpret most proportionally, with children instead relying on absolute quantity comparisons. However, recent research in proportional reasoning more generally has provided new insight into children’s apparent difficulties, revealing that their overreliance on absolute amount is unique to contexts in which the absolute amount can be counted and interferes with proportional information. Across two experiments, we test whether 4- to 6-year-old children’s interpretation of most is similarly dependent on the discreteness of the stimuli when comparing two different quantities (e.g., who ate most of their chocolate?) and when verifying whether a single amount can be described with the term most (e.g., is most of the butterfly colored in?). We find that children’s interpretation of most does depend on the stimulus format. When choosing between absolutely more vs. proportionally more as depicting most, children showed stronger absolute-based errors with discrete stimuli than continuous stimuli, and by 6-years-old were able to reason proportionally with continuous stimuli, despite still demonstrating strong absolute interference with discrete stimuli. In contrast, children’s yes/no judgements of single amounts, where conflicting absolute information is not a factor, showed a weaker understanding of most for continuous stimuli than for discrete stimuli. Together, these results suggest that children’s difficulty with most is more nuanced than previously understood: it depends on the format and availability of proportional vs. absolute amounts and develops substantially from 4- to 6-years-old.