Reduced oxidation of fat leading to a positive fat balance could be a factor in the development of obesity. Twenty-four-hour respiratory quotient (RQ) was measured in 152 nondiabetic Pima Indians fed a weight-maintenance diet [87 males and 65 females; 27 +/- 6 yr (mean +/- SD); 93.9 +/- 22.9 kg; 32 +/- 9% fat]. Twenty-four-hour RQ varied from 0.799 to 0.903. Prior change in body weight, 24-h energy balance, sex, and percent body fat explained 18% of the variance in 24-h RQ (P less than 0.001). In a subgroup of 66 siblings from 28 families, family membership explained 28% of the remaining variance in 24-h RQ (P less than 0.05). In 111 subjects for whom follow-up data (25 +/- 11 mo) were available, 24-h RQ was correlated with subsequent changes in body weight and fat mass (r = 0.27, P less than 0.01 and r = 0.19, P less than 0.05, respectively). Subjects with higher 24-h RQ (90th percentile) independent of 24-h energy expenditure were at 2.5 times higher risk of gaining greater than or equal to 5 kg body weight than those with lower 24-h RQ (10th percentile). We conclude that in Pima Indians fed a standard diet 1) family membership is the principal determinant of the ratio of fat to carbohydrate oxidation, and 2) a low ratio of fat to carbohydrate oxidation is associated with subsequent weight gain independent of low energy expenditure and may contribute to the familial aggregation of obesity.