OBJECTIVE: Promoting four healthy behaviours among preschool children:⩾4 servings of fruit and vegetables/day, ⩾2 h/day of active play, ⩽1 h per day of TV-watching and 0 sugar sweetened beverages/day. METHODS: We conducted a c-RCT on 425 3-year-old children at 16 childcare centres based in Cesena, Italy. We randomly allocated eight childcare centres (199 children) to the intervention group and eight childcare centres (226 children) to the control group. All the randomized childcare centres completed our study protocol. Parents recorded their children's target behaviours at home over 3 Saturdays, at baseline and at follow-up. Then trained nurses measured children's weight and height. We conducted a 6-month-long intervention trial in local health care centres where nurses and primary care pediatricians, respectively, conducted two subsequent motivational interviews with parents to encourage children's healthy behaviours at home. At the same time, teachers involved children in learning experiences about healthy behaviours. Our primary outcome is a children's combined health behaviour score (CHBS) at home. Our secondary outcomes measure the BMI z-score and the percentage of children that show a BMI trajectory crossing upward. After collecting the CHBS and BMI data at baseline as well as at 1- and 2-year follow-ups, we performed an Intent-to-Treat (ITT) analysis. RESULTS: After 2 years from baseline, 48.4% of intervention group children showed a low-risk CHBS in comparison with 28.0% of control group children. A multilevel analysis showed that they were by far more likely to achieve low-risk scores (adjusted OR: 3.41; 95% CI: 1.48-7.88; P=0.004). Our BMI outcomes showed no significant difference between groups. CONCLUSIONS: A multidimensional educational intervention, which consists of motivational interviews with parents and teacher-led learning experiences for children, improved preschool children's CHBS in the long term without influencing the outcomes of BMI z-score and BMI increase.