Aberrant crypt foci can be identified in the colons of rodents treated 3 wk earlier with azoxymethane, a known colon carcinogen. These crypts can easily be visualized in the unsectioned methylene blue-stained colons under light microscopy, where they are distinguished by their increased size, more prominent epithelial cells, and pericryptal space. They occur as single aberrant crypts or as two, three, or four aberrant crypts in a cluster. We compared the reported ability of carcinogens associated with the human diet to induce colon cancer with the measured rate of induction of aberrant crypts in female CF1 mice and Sprague-Dawley rats. The carcinogens used were 1,2-dimethylhydrazine, methyl nitrosourea, N-nitrosodimethylamine, benzo(a)pyrene, aflatoxin B1, 2-amino-6-methyldipyrido[1,2-alpha:3',2'-d]imidazole, 2-amino-3-methylimidazo[4,5-P]quinoline, 2-amino-3,4-dimethylimidazo[4,5-P]quinoline, and 3-amino-1-methyl-5H-pyrido[4,3-b]indole. Graded doses of these compounds were given to the animals by gavage twice with a 4-day interval, and the animals were terminated 3 wk later. All colon carcinogens induced aberrant crypts in a dose-related fashion. N-Nitrosodimethylamine and 3-amino-1-methyl-5H-pyrido[4,3-b]indole, carcinogenic compounds that do not induce colon cancer, did not induce them. The ability of the studied compounds to induce aberrant crypts was species specific; e.g., aflatoxin B1 and 2-amino-3,4-dimethylimidazo[4,5-P]quinoline induce about 20 times more in rats than mice. This relationship was consistent with their reported ability to induce colon cancer in these species. Results of the present study support the use of the aberrant crypt assays to screen colon-specific carcinogens and to study the process of colon carcinogenesis.