There is accumulating evidence suggesting that expression of genes for repair of UV damage to DNA in mammals and fish is regulated developmentally. Therefore, the activity of excision repair and photoreactivation in vivo in young larvae of Drosophila melanogaster was examined in a strain carrying the mutation mus201 that was unable to carry out excision repair. The photoreactivation activity in firstinstar larvae was so high that UV-induced lethality in excisionless larvae was almost completely rescued by posttreatment with fluorescent light. Excision repair activity in first-instar repair-proficient larvae was so high that UV irradiation was scarcely able to produce somatic eye-color mutations. In contrast, excisionless larvae showed a high incidence of somatic eye-color mutation after UV-irradiation, and this incidence was reduced to the spontaneous level by posttreatment with fluorescent light. Incorporation of a postreplication repair-defective mutation into the excisionless strain decreased the incidence of UV-induced somatic mutations by a factor of 3. The analogous repair dependence of UV mutagenesis in Drosophila and Escherichia coli is discussed. It is proposed that UVinduced somatic mutations in excisionless Drosophila larvae are caused primarily by pyrimidine dimers and that a constitutive, error-prone pathway for filling daughter-strand gaps opposite dimers is, at least partly, responsible for the fixation of mutations.The ability of living organisms to recover from lethal damage caused by external agents was first convincingly demonstrated in 1949 by Kelner (1) (4,8). Photoreactivation of UV-induced pyrimidine dimers occurs in the skin (only in the dermis) of neonatal mice but not in that of adult mice of the same inbred strain (9). In mouse embryo fibroblasts, excision-repair activity decreases abruptly during progressive passage of the cells in culture (10). Cultured mouse fibroblasts have only low activity of excision repair (11). In adult mice, repair in the dark of DNA damage caused by UV and UV-mimetic chemicals, as measured by unscheduled DNA synthesis in vivo in skin, is considerably less active in fibroblastic cells than in epithelial cells (12, 13). Dark repair activity for UV-induced DNA damage is absent in cultured fish cells (14,15) but present in fish embryos (15).The above cited findings strongly suggest that the expression of repair genes in animals is regulated developmentally (4). Much information is available on both the developmental biology and genetics of D. melanogaster. In this species, about 30 DNA repair-defective mutations have been reported (16), and two mutant strains, referred to by the mutations mei-9 and mus201 that they respectively carry, are unable to excise UV-induced pyrimidine dimers (16,17). Survival to eclosion of larvae of the mus201 strain is drastically reduced after UV-irradiation as compared with parallel irradiation of repair-proficient larvae (17). In the present study, we used this excisionless strain in studies on photoreactivation and dark ...