Objective: To evaluate the effectiveness of a nutritional intervention to promote fruit and vegetable (FV) intake. Design: A randomised controlled community trial was conducted to evaluate the effectiveness of a 7-month nutritional intervention and to promote FV intake, separately and together. All participants attended physical exercise sessions. The intervention was based on the transtheoretical model and Paulo Freire's pedagogy. The interventions included group educational sessions, motivational cards and informational materials. The primary outcome was a change in FV intake (g/d), and secondary outcomes included stages of change, self-efficacy, decisional balance and knowledge on FV. All data were collected face-to-face; and FV intake was assessed using a validated brief questionnaire. Setting: Health promotion services of Brazilian Primary Health Care. Participants: 3414 users of Brazilian Primary Health Care (1931 in the control group and 1483 in the intervention group (IG)). Results: At baseline, the average daily FV intake was 370 center dot 4 g/d (95 % CI 364 center dot 2, 376 center dot 6). The increase in FV intake (23 center dot 4 g/d; 95 % CI 6 center dot 7, 40 center dot 0) and fruit intake (+17 center dot 3 g/d; 95 % CI 5 center dot 1, 29 center dot 4; P = 0 center dot 01) was greater in the IG among participants in the lowest baseline intake. Participants in the IG also showed progression in the stages of change (P < 0 center dot 001), increased self-efficacy (P < 0 center dot 001) and improved knowledge of FV crops (P < 0 center dot 001). Conclusions: The nutritional intervention was effective in increasing FV intake and fruits intake among individuals with a lower intake at baseline and in maintaining FV intake among those who reported consuming FV as recommended (400 g/d).