This study was designed to determine the effect of local inflammation on nickel subsulfide (Ni3S2) carcinogenesis. Male F344/NCr rats, 6-week-old, 20 rats/group, received a single i.m. injection of 2.5 mg of Ni3S2 alone or 2.5 mg of Ni3S2 mixed with either 0.5 mg of Mycobacterium bovis lyophilized cell walls (MB), 1 mg cortisol (CORT), or 1 mg indomethacin (IND). Two more groups of rats received i.m. Ni3S2 alone, as above, followed by a single s.c. dose of either 1 mg MB or 2 mg IND. The i.m. injections were made into the thigh muscles of both hind limbs; the s.c. injections were made at the neck area. The experiment was terminated at 71 weeks. The final yields of injection site sarcomas were 85% in the Ni3S2, 85% in the Ni3S2 + CORT, and 80% in the Ni3S2 + IND group. Only one injection site tumor (5%) was found in rats given i.m. Ni3S2 + MB. In contrast, the s.c. MB injection enhanced muscular carcinogenicity of Ni3S2 by shortening the latency and increasing the yield of tumors to 100% at week 39 (P less than 0.04 vs. Ni3S2 alone). The s.c. treatment with IND gave similar, though statistically nonsignificant results; 100% tumor yield was reached at week 42 (P less than 0.18 vs. Ni3S2 alone). Although local treatment with CORT or IND had no significant effect on the final tumor incidence by Ni3S2, it shortened the latency of tumors to 17 weeks compared with 23 weeks for Ni3S2 alone. MB, CORT, IND, or the injection vehicle (water) alone did not produce tumors. Local MB treatment had no significant effect on the retention of Ni3S2 at the injection site. The prevention of the Ni3S2 tumors by local MB might result from localization of numerous natural killer cells and macrophages and formation of giant cells observed at the injection site of Ni3S2, 1-14 days post injection. These cells could immobilize the carcinogen and destroy Ni3S2-transformed cells.