OBJECTIVES: To identify and describe dietary patterns in a cohort of pregnant women and investigate whether the dietary patterns are associated with dietary intake and sociodemographic factors. DESIGN: Mothers entering the Finnish Type 1 Diabetes Prediction and Prevention (DIPP) Nutrition Study in 1997-2002 were retrospectively asked to complete a food-frequency questionnaire concerning their diet during pregnancy. Principal components analysis was used to identify dietary patterns. SETTING: Finland. SUBJECTS: Subjects were 3730 women with a newborn infant carrying increased genetic susceptibility to type 1 diabetes mellitus. RESULTS: Seven factors were identified and named. Energy intake correlated positively with 'Healthy', 'Fast food', 'Traditional bread', 'Traditional meat' and 'Coffee' patterns and inversely with the 'Alcohol and butter' pattern. Intake of dietary fibre correlated positively with 'Healthy', 'Traditional bread' and 'Low-fat foods' patterns and inversely with the 'Alcohol and butter' pattern. The seven dietary patterns seemed to account for relatively large proportions of the variance in energy and nutrient intakes except for the intake of vitamin D, vitamin C, carotenoids and calcium. Maternal age and higher level of education were associated with higher scores on 'Healthy', 'Low-fat foods' and 'Alcohol and butter' patterns. CONCLUSION: Principal components analysis produced seven dietary patterns which may be useful for further research concerning maternal diet and health outcomes among both mothers and their offspring.