Evaluating in vivo photochemical genotoxicity (photogenotoxicity) or photochemical carcinogenicity (photocarcinogenicity) in the skin that is actually exposed to light is important for estimating the risk of human exposure to chemicals under sunlight. With regard to the skin micronucleus test, Nishikawa et al. developed a reliable technique that is simple and in which the negative control has a stable background. In the present study, we applied 8-methoxypsoralen (8-MOP) and benzo[a]pyrene (B[a]P) to the backs of hairless mice and subjected the mice to irradiation by a sunlight simulator in order to investigate whether this test can detect photogenotoxicity of these chemicals. In the treatment with 8-MOP [0.00075% and 0.0015% (w/v)], a significant increase was observed in the frequency of micronucleated cells only under light irradiation using the sunlight simulator. At a high chemical dose, the frequency of micronucleated cells increased from 48h after the treatment, peaked at 96h, and then decreased at 168h. Furthermore, at 96h with the high dose under light irradiation, we frequently observed cells with nuclear buds. In the treatment with B[a]P [first experiment: 0.025% and 0.05% (w/v); second experiment: 0.005%, 0.01%, and 0.02% (w/v)], a significant increase was observed in the frequency of micronucleated cells at skin-irritating doses [0.01%, 0.02%, 0.025%, and 0.05% (w/v)] at 72 or 96h after the treatment only under light irradiation using the sunlight simulator. In conclusion, photogenotoxicity of 8-MOP and B[a]P was detected in the in vivo photochemical skin micronucleus study.