The dose dependencies of the lung carcinogenicity of 1,6-dinitropyrene (1,6-DNP) and benzo[a]pyrene (BaP) were examined by direct injections of these compounds into rat lungs. A total of 276 male F344 rats were divided into 10 groups and given various doses of 1,6-DNP or BaP, or no drug (control group). Both chemicals were injected into the lung, as suspensions in beeswax--tricaprylin and the animals were then observed for 104 weeks. The incidences of lung cancer were 0/39 (0%), 4/30 (13%), 13/31 (42%), 22/26 (85%) and 6/9 (67%) in groups treated with 0.003, 0.01, 0.03, 0.1 and 0.15 mg of 1,6-DNP respectively, and 1/29 (3%), 7/30 (23%), 22/29 (76%) and 9/13 (69%) in those treated with 0.03, 0.1, 0.3 and 1.0 mg of BaP respectively. No lung cancer was found in control rats. Thus the incidences of lung cancer induced by 1,6-DNP and BaP showed significant dose dependence. At equal doses, the incidence of lung cancer was much higher with 1,6-DNP than with BaP, and the induction of cancer by 1,6-DNP was higher even at one-third the dose of BaP. Histologically, most tumors induced by 1,6-DNP were undifferentiated neoplasms, whereas most of those induced by BaP were well-differentiated squamous cell carcinomas.