We compared changes in the dietary patterns of morbidly obese patients undergoing either laparoscopic gastric bypass surgery or a comprehensive lifestyle intervention programme. The present 1-year non-randomised controlled trial included fifty-four patients in the lifestyle group and seventy-two in the surgery group. Dietary intake was assessed by a validated FFQ. ANCOVA was used to adjust for between-group differences in sex, age, baseline BMI and baseline values of the dependent variables. Intakes of food groups and nutrients did not differ significantly between the intervention groups at baseline. At 1-year follow-up, the lifestyle group had a significantly higher daily intake of fruits and vegetables (561 (sd 198) v. 441 (sd 213) g, P= 0·002), whole grains (63 (sd 24) v. 49 (sd 16) g, P< 0·001) and fibre (28 (sd 6) v. 22 (sd 6) g, P< 0·001) than the surgery group and a lower percentage of total energy intake of saturated fat (12 (sd 3) v. 14 (sd 3) %, P< 0·001). The intake of red meat declined significantly within both groups, vegetables and fish intake were reduced significantly in the surgery group and added sugar was reduced significantly in the lifestyle group. The lifestyle patients improved their dietary patterns significantly (compared with the surgery group), increasing their intake of vegetables, whole grains and fibre and reducing their percentage intake of saturated fat (ANCOVA, all P< 0·001). In conclusion, lifestyle intervention was associated with more favourable dietary 1-year changes than gastric bypass surgery in morbidly obese patients, as measured by intake of vegetables, whole grains, fibre and saturated fat.