The present study was designed to establish the form of the dose-response relationship for dietary sodium arsenite as a co-carcinogen with ultraviolet radiation (UVR) in a mouse skin model. Hairless mice (strain Skh1) were fed sodium arsenite continuously in drinking water starting at 21 days of age at concentrations of 0.0, 1.25, 2.5, 5.0, and 10 mg/L. At 42 days of age, solar spectrum UVR exposures were applied three times weekly to the dorsal skin at 1.0 kJ/m2 per exposure until the experiment ended at 182 days. Untreated mice and mice fed only arsenite developed no tumors. In the remaining groups a total of 322 locally invasive squamous carcinomas occurred. The carcinoma yield in mice exposed only to UVR was 2.4 +/- 0.5 cancers/mouse at 182 days. Dietary arsenite markedly enhanced the UVR-induced cancer yield in a pattern consistent with linearity up to a peak of 11.1 +/- 1.0 cancers/mouse at 5.0 mg/L arsenite, representing a peak enhancement ratio of 4.63 +/- 1.05. A decline occurred to 6.8 +/- 0.8 cancers/mouse at 10.0 mg/L arsenite. New cancer rates exhibited a consistent-with-linear dependence on time beginning after initial cancer-free intervals ranging between 88 and 95 days. Epidermal hyperplasia was elevated by arsenite alone and UVR alone and was greater than additive for the combined exposures as were growth rates of the cancers. These results demonstrate the usefulness of a new animal model for studying the carcinogenic action of dietary arsenite on skin exposed to UVR and should contribute to understanding how to make use of animal data for assessment of human cancer risks in tissues exposed to mixtures of carcinogens and cancer-enhancing agents.