The aim of this study is to examine the effects of growth hormone, exercise, and weight loss due to food restriction on tibial diaphyseal bone and on tibial muscle mass. Thirteen-month-old female F344 rats were divided into six groups: group 1, baseline controls (B); group 2, age-matched controls (C); group 3, GH treated (GH); group 4, voluntary wheel running exercise (EX); group 5, GH + EX; and group 6, food restricted (FR). The dose of GH was 2.5 mg recombinant human (rh) GH/kg body weight/day, 5 days per week, given in two divided doses of 1.25 mg at 9-10 A.M. and 4-5 P.M. Food-restricted rats were fed 60% of the mean food intake of the age-matched controls. All animals except the baseline controls were killed after 4.5 months. The baseline controls were killed at the beginning of the study. Growth hormone increased the body weight and tibial muscle mass of the rats markedly, while EX caused only a slight decrease in body weight and partially inhibited the increase caused by GH in the GH + EX group. Food restriction greatly decreased body weight below that of age-matched controls, but neither FR nor EX had a significant effect on the mass of the muscles around the tibia. Growth hormone and EX independently increased tibial diaphyseal cortical bone area (p < 0.0001, p < 0.0001), cortical thickness (p < 0.0001, p < 0.0001), cortical bone mineral content (p < 0.0001, p < 0.0001), periosteal perimeter (p < 0.0001, p < 0.0001), and bone strength-strain index (SSI) (p < 0.0001, p < 0.0001). The effects of GH were more marked and resulted in a greater increase in the weight of the mid tibial diaphysis (p < 0.0001). The combination of GH and EX produced additive effects on many of the tibial diaphyseal parameters, including bone SSI. GH + EX, but not GH or EX alone, caused a significant increase in endocortical perimeter (p < 0.0001). In the FR rats, cortical bone area and cortical mineral content increased above the baseline level (p < 0.001, p < 0.0001) but were below the levels for age-matched controls (p < 0.0001, p < 0.0001). In addition, marrow area, endocortical perimeter, and endocortical bone formation rate increased significantly in the FR rats (p < 0.01, p < 0.0001, p < 0.0001). Three-point bending test of right tibial diaphysis resulted in maximum force (Fmax) values that reflected the group differences in indices of tibial diaphyseal bone mass, except that GH + EX did not produce additive effect on Fmax. The latter showed good correlation with left tibial diaphyseal SSI (r = 0.857, p < 0.0001), and both indices of bone strength correlated well with tibial muscle mass (r = 0.771, Fmax; r = 0.700, SSI; p < 0.0001). GH increased serum IGF-I (p < 0.0001), and the increase was partially reduced by EX. Serum osteocalcin was increased by GH with or without EX (p < 0.01, p < 0.01), and FR or EX alone did not alter serum IGF-I and osteocalcin levels. The bone anabolic effects of GH with or without EX may relate, in part, to increased load on bone from tibial muscles and body weight, which were increased by the hormone. The osteogenic effect of EX with or without GH may relate, in part, to increased frequency of muscle load on bone as EX decreased body weight (p < 0.05), but had no significant effect on tibial muscle mass. The enhanced loss of endocortical bone by FR may relate, in part, to decreased load on bone due to low body weight (p < 0.0001), as FR did not cause a significant decrease in tibial muscle mass (p = 0.357). The roles of humoral and local factors in the bone changes observed remain to be established.