OBJECTIVES: Analysing data of the Young Lives Study in Peru, we aimed at assessing the association between daily food frequency and body mass index (BMI) changes between 2006 and 2016. DESIGN: Secondary analysis of a prospective ongoing cohort study. SETTING: 20 sentinel sites in Peru. PARTICIPANTS: Children enrolled in the younger cohort of the Young Lives Study. We used information from the second (2006-2007), third (2009-2010), fourth (2013-2014) and fifth (2016-2017) rounds of the younger cohort in Peru. PRIMARY AND SECONDARY OUTCOMES: BMI as well as BMI-for-age z-score, both as numerical variables. RESULTS: Data from 1948 children, mean age 4.3 (SD: 0.3) years and 966 (49.6%) women were included at baseline. In multivariable model, lower food consumption frequency was associated with increased BMI and BMI-for-age z-scores: children reporting <4 times of food consumption per day had a greater increase in BMI (β=0.39; 95% CI 0.17 to 0.62) and BMI-for-age z-score (β=0.07; 95% CI 0.01 to 0.13) compared with those reporting 5 per day. Results were consistent for those reporting exactly eating 4 times per day (β for BMI=0.16; 95% CI 0.02 to 0.30 and β for BMI-for-age z-score=0.05; 95% CI 0.01 to 0.09). CONCLUSIONS: Children who eat <5 times per day, gain more BMI compared with those who eat ≥5 times. Parents should receive information to secure adequate nutrition for their children, both in terms of quality and quantity.