The cocarcinogenic effect of chrysotile asbestos was investigated in heterotopic tracheal transplants of F344 rats. Tracheal transplants were first exposed to graded doses of dimethylbenz[a]anthracene (DMBA) contained in intraluminal pellets. The doses ranged from 12.5 to 100 micrograms. Control tracheas received blank pellets. Four weeks after the start of DMBA exposure (when all carcinogen had been released from the pellets), the spent pellets were removed, and 200 micrograms chrysotile, a nontumorigenic dose of asbestos, was introduced into the lumina of the preexposed tracheas. No significant enhancement of the tumor response was seen with 100 micrograms DMBA, a dose that was tumorigenic by itself. However, with 50 and 25 micrograms DMBA, nontumorigenic dose levels, a 15 and 23% incidence of tracheal carcinomas occurred when DMBA exposure was followed by a nontumorigenic dose of chrysotile. At 12.5 micrograms DMBA, this effect was not observed. Whatever the mechanisms were that formed the basis of this tumor enhancement effect of asbestos, the consequences were similar to those observed with tumor promotion in classical two-stage carcinogenesis studies.