OBJECTIVE: To investigate the relationships among subcutaneous fatness, subcutaneous adipose tissue (SAT) distribution, somatotype and risk factors for coronary heart disease (CHD). SUBJECTS: The sample included 1410 (715 male and 695 female) youths and adults from the Que  bec Family Study. MEASUREMENTS: Six skinfolds and the dimensions necessary for the derivation of the Heath-Carter anthropometric somatotype (endomorphy, mesomorphy, ectomorphy) were measured. The six skinfolds were summed to provide an index of subcutaneous adiposity (SUM). In addition, the trunk-to-extremity skinfold ratio, adjusted for SUM using regression procedures (TER), and the ®rst principal component (PC1) of skinfold residuals (also adjusted for SUM) were used to indicate SAT distribution, independent of the overall level of fatness. Risk factors for CHD included systolic and diastolic blood pressures, and fasting glycaemia, triglycerides (TGs), plasma cholesterol, high and low density lipoprotein (HDL-C and LDL-C) cholesterol, and the HDL-Catotal cholesterol (CHOL) ratio. RESULTS: In general, SUM was positively correlated with endomorphy and mesomorphy, and negatively correlated with ectomorphy. On the other hand, SAT distribution was not associated with somatotype, except in females where TER and PC1 were negatively correlated with mesomorphy. Results of forward stepwise regression analyses to predict CHD risk factors, indicated that a signi®cant proportion of the variance in the risk factors could be accounted for by SUM, SAT distribution and somatotype (up to 16%). SUM is the best predictor, entering the regressions ®rst (most important) in six of 15 signi®cant regressions in males and 14 of 16 signi®cant regressions in females. Somatotype components enter as predictors 10 times in males, and six times in females. Similarly, TER and PC1 enter as predictors nine times in males and ®ve times in females. CONCLUSIONS: Somatotype is related to SUM, while somatotype and SAT distribution are largely independent of one another. Furthermore, SUM, somatotype and SAT distribution are signi®cant predictors of biological risk factors for CHD.