The incidence and latency periods for local tumor development after sc injection of 3-methylcholanthrene (MCA) into 30-day-old nude mice (nu/nu partially inbred on the CBA/H background) derived from homozygous matings (nu/nu times nu/nu) or heterozygous matings (nu/+ times nu/+) were comparable and did not differ with the immunologically normal controls, even when the carcinogen dosages ranged from 0.01 to 0.10 mg. Similarly, no differences in tumor incidence or latency periods between nude mice from homozygous or heterozygous matings as well as their immunologically normal controls were observed when weight-adjusted doses of MCA equivalent to 0.02 to 0.10 mg in the 30-day-old mice were administered at 120, 210, or 360 days of age. Tumor incidence was lower in nude mice and normal mice when MCA was administered at 210 and 360 days of age, especially in mice given the lower dose of MCA. The lower dosages of MCA (0.01-0.05 mg) had no detectable immunodepressive effects in normal mice. Thus the "normal" tumor incidence in nude mice after MCA administration could not be attributed to: 1) the effect of humoral thymus gland function (in the nude mice derived from heterozygous matings), 2) the immunodepressive effects of the carcinogen (the lower MCA dosages are not immunodepressive), or 3) the age of the mice at administration. These results argue against the thymus dependency of immunologic surveillance.