Although previous studies demonstrated that alcohol consumption is associated with low chronic kidney disease (CKD) risk, they did not consider individual body mass and metabolic capacity. We examined whether the body mass index (BMI) affects the association between drinking and CKD. We defined CKD as an estimated glomerular filtration rate decline < 60 mL/min/1.73 m2 and/or dipstick proteinuria (≥ 1+). Participants were 11,175 Japanese individuals aged 40–74 years without baseline CKD who underwent annual health checkups. Daily alcohol consumption was estimated from a questionnaire, and the participants were categorized as "infrequent drinkers," "light drinkers (< 20 g/day)," "moderate drinkers (20–40 g/day)," and "heavy drinkers (≥ 40 g/day)." Over a median 5-year observation period, 936 participants developed CKD. Cox proportional hazards models revealed that light drinkers had a significantly reduced risk of CKD compared with infrequent drinkers (P = 0.01). Stratified by BMI, the low BMI (< 18.5 kg/m2) group had an increased risk of CKD even in light drinkers, while the high BMI (≥ 25.0 kg/m2) group had a decreased risk of CKD regardless of alcohol consumption. Taken together, alcohol consumption did not reduce the CKD risk in all populations; individual tolerance must be considered.