BACKGROUND: The aim was to investigate whether social network and social support factors can explain socioeconomic differences in the risk of consuming low amounts of vegetables, fruit and fruit juices. METHODS: The Malm? Diet and Cancer Study was a prospective cohort study. The present cross-sectional study examined data from a subpopulation of 11,837 individuals that completed baseline examinations in 1992-1994. Dietary habits were assessed using a modified diet history method, and socioeconomic and social network factors were measured with a structured questionnaire. Low consumption was defined as the lowest consumption quartile for vegetables and fruit, while fruit juice consumption was dichotomized to separate users from non-users. RESULTS: Socioeconomic differences were most pronounced regarding the consumption of vegetables and fruit juices. For both sexes, unskilled manual workers had a twice as high risk of low vegetable and fruit juice consumption as higher non-manual employees. No socioeconomic differences in fruit consumption were observed for men, and only moderate differences for women with a higher consumption in higher socioeconomic groups. When the psychosocial variables were introduced in the multivariate model, social participation moderately reduced the socioeconomic differences in vegetable consumption, and the female socioeconomic differences in fruit consumption, but had no effect on the socioeconomic differences in fruit juice consumption. The other psychosocial variables had no effect on the socioeconomic differences. CONCLUSION: Considerable socioeconomic differences in vegetable, fruit and fruit juice consumption were observed. Social participation seemed to be a strong determinant for these food choices. However, this effect was largely independent of the socioeconomic differences.