DNA is taken as the target for skin cancer induced by ultraviolet light, and the known data on the sensitivity of DNA as a function of wavelength are summarized. The sun's spectrum at the surface of the earth and the DNA action spectrum are used to calculate the carcinogenic effectiveness as a function of wavelength.The most effective wavelengths at 30'N latitude are <305 nm, and a 1% change in atmospheric ozone results in a 2% change in the effective dose of ultraviolet light. Since both the basic biological and physical data are reasonably precise, the major requirement for a quantitative evaluation of the dose response relation for ultraviolet-induced skin cancer in man is better epidemiological data to compare with data from animal models.Human skin cancers, especially basal and squamous cell carcinomas, are closely associated with exposure to sunlight.[See reviews by Blum (1), Epstein (2), and the volume edited by Urbach (3). ] Three lines of evidence indicate that the most effective wavelengths are below 320 nm. (i) In mice, wavelengths longer than 320 nm are ineffective in inducing skin cancer although a recent report indicates that exposure to the longer wavelengths may accentuate the effects of shorter ones (4). (ii) The effective wavelengths for erythema production are below 320 nm. Skin cancer and erythema arise in the same tissue, and individuals who sunburn easily have a higher probability than average of developing skin cancer (5, 6). (iii) Ultraviolet light (UV)-induced skin cancer probably arises from photochemical changes in DNA, and the shorter wavelengths are much more effective than. the longer ones in damaging this polymer (see below). Interest in wavelength dependence arises not only because of the inherent interest in the problem but also for the practical reason of estimating the effects of this model environmental hazard. The hazard could change. For example, the exhausts from a fleet of supersonic transports might result in a decrease in stratospheric -ozone and the attendant increase in UV fluence at the earth's surface could result in an increase in the incidence rate of skin cancer. This particular problem has been succinctly stated (7). If we are to evaluate quantitatively such hazards, we must have good animal and epidemiological data as well as a theoretical framework to handle such data.UV-induced skin cancer deserves more careful epidemiological study because (i) we know more about UV-induced lesions in DNA than any other physicochemical insult to the Abbreviation: UV, ultraviolet light.