Composition of diets may influence growth, diseases, tumor rates, and responses to chemical treatment. For two years, Fischer 344 rats were fed the NIH-07 open-formula nonpurified diet (approximately 23% protein, approximately 5% fat, and approximately 3.5% fiber) and nonpurified experimental diets (NTP-90, NTP-91, and NTP-92) containing lower protein and higher fat and fiber (14.6-15.3% protein, 7.2-8.5% fat, and 9.4-14% fiber) than the NIH-07 diet. Rats were evaluated for growth patterns, survival, hematology, serum chemistry, nephropathy, and tumor incidences. Growth patterns were similar in rats fed the experimental diets and in those fed the NIH-07 diet. However, in rats fed the experimental diets, the adult body weights were significantly (6-9%) lower and the survival at 110 weeks of age was significantly higher (15-20%) than in rats fed the NIH-07 diet. Lower protein content of experimental diets decreased the severity of nephropathy. Higher fat content of experimental diets appears to have decreased the incidence or delayed the development of leukemia and associated mortality in males. Higher fiber content of experimental diets appears to have delayed the development of mammary tumors and associated mortality in females. Higher fat and/or fiber of the experimental diets decreased the incidence of pheochromocytomas in males. The lower protein and higher fat and fiber contents of the experimental diets decreased the spontaneous tumor burden in two-year studies. These studies indicate that diets for rats in long-term studies could be modified to decrease the severity of nephropathy and to decrease/delay the development of spontaneous tumors.