Objective: To investigate the cross-sectional association between food intake and overweight in children. Methods: Height and weight were measured in 2,565 school-aged children. Intakes of 11 food groups were categorized (low, medium and high) using specific tertile cutoffs. Multivariate energy partition models were applied. Adjustment included energy intake from other food groups, city, family income, parental education and 'screen' time. Possible underreporters were identified and used in sensitivity analyses. Results: Compared to low intake, high intakes of meat, fish, beverages and bakery products were associated with greater BMI z-scores [β (95% CI) = 0.32 (0.21, 0.42), 0.13 (0.03, 0.24), 0.23 (0.11, 0.35) and 0.10 (-0.01, 0.20)] and increased risk of being overweight [odds ratio (OR) (95% CI) = 2.08 (1.58, 2.73), 1.39 (1.08, 1.80), 1.36 (1.01, 1.84) and 1.62 (1.24, 2.11)]. Conversely, medium and high intakes of confectionery were associated with smaller BMI z-scores [β = -0.18 (-0.28, -0.07) and -0.22 (-0.33, -0.12)] and decreased risk of being overweight [OR = 0.64 (0.50, 0.83) and 0.53 (0.40, 0.68)]. These associations were robust to sensitivity analyses. Conclusions: Intakes of meat, fish, beverages and bakery products correlate with body weight status.