Pediatric obesity develops when a complex biological predisposition collides with an obesogenic environment. To further elucidate the role of genetics in obesity onset, we performed a candidate-gene association study in a young and sportive Italian population by testing the association of functional polymorphisms in ACE (rs4646994), FTO (rs9939609), MC4R (rs17782313) and PPARG (rs1801282) genes with body mass index (BMI) and waist-to-height ratio (WHtR). We also tested the combinations of identified risk genotypes and epistatic interactions among them to determine the existence of cumulative effects in predicting the predisposition to gain weight. Our results confirm a significant direct influence of MC4R rs17782313 and PPARG rs1801282 on body composition, that is, minor allele homozygotes showed significantly higher BMI (rs17782313, β = 1.258, P = 0.031; rs1801282, β = 6.689, P = 1.2 × 10 ) and WHtR (rs17782313, β = 0.021, P = 0.005; rs1801282, β = 0.069, P = 0.003) values. Moreover, by leveraging multifactor dimensionality reduction and general linear model (GLM) approaches we identified an epistatic interaction between ACE and MC4R, where heterozygosity at ACE rs4646994 seems to protect from the unfavorable predisposition to gain weight given by C/C genotype at MC4R rs17782313 (GLM, P = 0.004). In conclusion, to clarify the role of genetics in multifactorial diseases remains a difficult goal, even for the most investigated polymorphisms and in controlled populations. Further studies on epistasis and gene-gene interaction will help to elucidate this complex scenario. © 2017 IUBMB Life, 69(2):98-105, 2017.