Maximal static, eccentric and concentric torques and arm components estimated by anthropometry and measured by computed tomography were evaluated in 25 male monozygotic twins and 16 dizygotic twins (22.4 +/- 3.7 years). The importance of genetic and environmental factors in the observed variation in these measurements was estimated by genetic model-fitting techniques. In this sample of young adult male twins, genetic factors were significant in most of the strength measurements, arm muscle components and muscle activation variables. The contribution of genetic factors in strength measures depended on the angle, contraction type and to some extent on contraction velocity. For isometric strength, angle-specificity in genetic and environmental variation could be attributed to the degree of variability in muscle activation and performance discomfort at each specific angle, with the highest unique environmental impact at extreme angles. The high genetic contribution at 170 degrees, but not at 50 degrees, possibly expressed different contributions of genetic factors in the muscle-length factor and moment arm in torques at both angles. The importance of genetic factors in eccentric arm flexor strength (62-82%) was larger than for concentric flexion (29-65%), as the pattern of genetic determination followed the torque-velocity curve. Genetic variations in contractile and elastic components, contributing differently to eccentric and concentric torques, together with velocity-dependent actin-myosin binding factors, could account for the observed differences. The broad heritability was very high for all anthropometric and arm cross-sectional area measurements (> 85%) and common environmental factors were only significant for anthropometrically estimated mid-arm muscle tissue are (48%). Heritability estimates of different arm muscularity measurements were comparable.