This study was designed to determine the effects of four well-characterized dietary brans on large bowel tumorigenesis induced in mice with 1,2-dimethylhydrazine (DMH). Eight-week-old barrier-derived male Balb/c mice were fed a semisynthetic diet with 20% bran added (either corn, soybean, soft winter wheat, or hard spring wheat) or a no-fiber-added control diet. Half of each group was given DMH (20 mg/kg body weight/week, subcutaneously for 10 weeks) beginning at 11 weeks of age. Surviving mice were killed 40 weeks after the first DMH injection. Tumors were not found in mice not subjected to DMH. In DMH-treated mice, tumors were found almost exclusively in the distal colon. Tumor incidences were as follows: controls, 11%; soybean group, 44%; soft winter wheat group, 48%; hard spring wheat group, 58%; and corn group, 72%. Tumors per tumor-bearing mouse ranged from 1.4 to 1.6, except in the corn group, which had 2.1. A positive correlation was found between percentage of neutral detergent fiber in the brans and tumor incidences but not between the individual components of cellulose, hemicellulose, or lignin. The enhancement of DMH-induced large bowel tumorigenesis by all four bran types may reflect a species and/or mouse strain effect that is bran-source related. These data emphasize the importance of using well-defined bran in all "fiber" studies.