The function and morphology of knee extension/m. vastus lateralis and elbow flexion/m. biceps brachii were studied in young (28 +/- 0.1 years, n = 7) and elderly (68 +/- 0.5 years, n = 8) sedentary subjects and in elderly swimmers (69 +/- 1.9 years, n = 6), runners (70 +/- 0.7 years, n = 5) and strength-trained subjects (68 +/- 0.8 years, n = 7). On average, the training groups had, for the 12-17 years before the measurements were taken, performed their training regimen 3 +/- 0.1 times a week. Compared with the young subjects, the maximal isometric torque of the sedentary elderly subjects was 44% (P less than 0.05) lower in knee extension and 32% (P less than 0.05) lower in elbow flexion, and speed of movement was between 20 and 26% (P less than 0.05) lower in both knee extension and elbow flexion. The cross-sectional area of m. quadriceps femoris and the elbow flexors was also 24% (P less than 0.05) and 20% lower respectively, and the specific tension was 27% (P less than 0.05) lower in m. quadriceps femoris and 14% (P less than 0.05) lower in the elbow flexors. A 27% (P less than 0.05) higher content of myosin heavy chain type I and a 39% (P less than 0.05) higher content of the slow-type myosin light chain--2 was observed in m. vastus lateralis of the sedentary elderly subjects as compared with the young subjects. The same tendency was also seen with m. biceps brachii. Since the histochemical fibre-type distribution was identical and no major co-expression of type I and type II myosin heavy-chain isoforms was observed with immunocytochemistry, the increase in slow myosin isoforms with ageing seems mainly related to a larger relative area of type I fibres, induced by a selective atrophy of type II fibre area. An increased content of the beta-isoform of tropomyosin was also demonstrated with ageing. In contrast to the swimmers and runners, the elderly strength-trained subjects had maximal isometric torques, speed of movements, cross-sectional areas, specific tensions and a content of myosin and tropomyosin isoforms in both muscles studied identical to those of the young controls. These results seem to suggest that strength training can counteract the age-related changes in function and morphology of the ageing human skeletal muscle.