Moderate heritability for skeletal muscle strength has been reported in twin studies, but genetic co-variation between muscle strength at different parts of body and body size is not well known. Further, representativeness of twin cohorts needs to be critically evaluated. Height, weight, elbow flexion, hand grip and knee extension strength were measured in young adulthood in 1,139,963 Swedish men born between 1951 and 1976. We identified 154,970 full-brother pairs and 1582 monozygotic (MZ) and 1864 same-sex dizygotic (DZ) complete twin pairs. The data were analyzed using quantitative genetic modeling for twin and family data. Twins compared to singletons and MZ twins compared to DZ twins were shorter, lighter and had lower muscle strength. In singletons, there was more variation in weight and the strength measures compared to twins with known zygosity but not when compared to twins with unknown zygosity. Full-sib correlations for these traits were lower than DZ correlations. Additive genetic factors explained 81% of variation in height, 59% in body mass index and 50–60% in the strength measures. Additive genetic correlations varied from 0.13 between height and elbow flexion strength to 0.78 between elbow flexion and hand grip strength. Our results suggest that extra variation may exist in general populations not found in twin samples, probably because of selection due to non-participation. This may have inflated heritability estimates in previous twin studies. Nonetheless, we showed that genetic factors affect muscle strength and part of these genes are common to different strength indicators and body size. Genet. Epidemiol. © 2008 Wiley-Liss, Inc.