Objective: To describe changes in consumption of different types of beverages from pre-pregnancy to early pregnancy, and to examine associations with maternal age, educational level and BMI. Design: Cross-sectional design. Participants answered an FFQ at inclusion into a randomized controlled trial, the Fit for Delivery (FFD) trial, in median gestational week 15 (range: 9-20), reporting current consumption and in retrospect how often they drank the different beverages pre-pregnancy. Setting: Eight local antenatal clinics in southern Norway from September 2009 to February 2013. Subjects: Five hundred and seventy-five healthy pregnant nulliparous women. Results: Pre-pregnancy, 27 % reported drinking alcohol at least once weekly, compared with none in early pregnancy (P < 0·001). The percentage of women drinking coffee (38 % v. 10 %, P < 0·001), sugar-sweetened beverages (10 % v. 6 %, P = 0·011) and artificially sweetened beverages (12 % v. 9 %, P = 0·001) at least daily decreased significantly from pre-pregnancy to early pregnancy, while the percentage of women who reported to drink water (85 % v. 92 %, P < 0·001), fruit juice (14 % v. 20 %, P = 0·001) and milk (37 % v. 42 %, P = 0·001) at least daily increased significantly. From pre-pregnancy to early pregnancy higher educated women reduced their consumption frequency of coffee significantly more than women with lower education. Older women reduced their consumption frequency of coffee and artificially sweetened beverages and increased their consumption frequency of fruit juice and milk significantly more than younger women. Conclusions: There is a significant change in beverage consumption from pre-pregnancy to early pregnancy among Norwegian nulliparous women.