Studies were carried out to examine the effects of long-term recombinant human growth hormone (GH) therapy on longevity in rodents. In the first study, 150 18-month-old female F344 rats were divided into three groups of 50 rats per group: Group 1, solvent vehicle; Group 2, 10 microg GH/kg body weight three times per week; Group 3, 50 microg GH/kg body weight three times per week. GH and solvent vehicle therapies were started at 18 months of age and continued until all the animals died spontaneously. Serum insulin-like growth factor (IGF)-I was measured at 18 and 29 months of age and on 3-month-old rats. Serum IGF-I level decreased between 3 and 29 months of age. GH therapy reversed the decrease in a dose-dependent manner, with the 50 microg GH dose returning the serum IGF-I level to that of 3-month-old animals. However, statistical analysis revealed no significant effect of GH therapy on median life span, 10th percentile life span, or maximum life span. Similar observations on longevity were made on aged F344 male rats and on aged Balb/c mice, even when the dose of GH was increased to 1.0 mg/kg body weight two times per week. The main pathologic lesions in control animals were nephropathy, cardiomyopathy, leukemia, and testicular interstitial cell tumor; the prevalence of these lesions was not significantly altered by GH therapy. We conclude that long-term low-dose GH therapy that includes doses in the range that is given to humans in clinical trials in GH deficiency and to revert age-related physiologic declines has no overt deleterious effects on longevity and pathology in aged rodents.