The renal toxicity of harman and norharman, administered for 2 or 4 weeks at dietary levels of 1,000, 500, or 0 parts per million (ppm), was investigated in 6-week-old male F344/DuCrj rats. Although rats fed 1,000 ppm harman or norharman, but not the 500 ppm level, demonstrated marked body weight retardation from 1 week to termination, no mortalities occurred. Marked elevation of water consumption was evident in rats given harman or norharman at 1,000 ppm, but not at 500 ppm, together with large increases in urine of low specific gravity. Urinary lysosomal enzymes (N-acetyl-beta-D-glucosaminidase, NAG, and lactate dehydrogenase, LDH) and sugar levels were increased, and the brush border enzymes (gamma-glutamyl transpeptidase, GGT, and alkaline phosphatase, ALP) decreased. Furthermore, serum biochemistry revealed clear elevation of parameters indicating renal toxicity in these rats. Histopathologically, rats fed 1,000 ppm harman or norharman, but not 500 ppm, demonstrated focal toxic renal degenerative/necrotic and regenerative lesions in proximal, distal, and collecting tubules. These changes were associated with a clearly increased labeling index (LI) of the nuclei of renal tubular epithelial cells on immunohistochemical staining for 5-bromo-2'-deoxyuridine (BrdU). Chemical specific crystal formation within tubular lumina was evident in rats fed 1,000 ppm, but not 500 ppm, this being considered the cause of the renal tubular lesions. It was concluded that harman and norharman exert renal toxicity at the dietary level of 1,000 ppm, but not 500 ppm, in male F344 rats.