Purpose The purpose of this paper is to explore fruit and vegetable (FV) procurement disparity across income groups. Design/methodology/approach This study uses mean comparison and quintile regression to explain FVs variations. Findings Households from the highest income quantile spend more than two times on FVs than households from the lowest quantile; however, this expenditure disparity is largely mitigated in terms of purchase quantity. This paper presents evidence that, rather than quantity discounts or income neighborhood, the type of store (traditional markets vs supermarkets) plays a relevant role in explaining the smaller gap in terms of purchase quantity. Research limitations/implications Traditional markets help low-income households access low-cost FVs. Social implications The authors generate evidence to show that traditional markets play a relevant role to supply affordable FV to low-income households. Originality/value The paper used a high-quality and uncommon data set. It is a topic of high social impact.