This work is an exploration of some ways in which the physical characteristics of sunlight exert influence on the occurrence of sunburn and skin cancer. A summary of evidence relating skin cancer to radiation exposure includes published conclusions from Northern Hemisphere experience together with newer instances appearing in the Australian tropical population. A variety of evidence suggests that a study of skin cancer may be based on the assumption that the carcinogenic action spectrum is approximately the same as for sunburn independently of the fact that frequent sunburn accelerates the appearance of skin tumours. A very brief and somewhat speculative review is presented of biological aspects of skin cancer which are clearly related to radiation exposure. Variations in the relative effectiveness of sunlight have been calculated by combining the solar intensity spectrum at sea-level with the action spectrum for erythema (mild sunburn) of the skin. This is presented to show the influence of response spectrum, solar altitude, latitude, time of year and time of day.Methods of measurement are discussed by which an instrumental record may be obtained representing the erythemal or carcinogenic effectiveness of sunlight on human skin. Development and use of a fluorescent system is described in which magnesium tungstate phosphor is used to evaluate the different photon energies with a relative spectral responsiveness which is quantitatively similar to that for skin. Associated arrangements for measuring the resulting small currents are also included.Continuous records, covering several years, of the erythemal effectiveness of sunlight have been obtained by measurement at 7 sites ranging from 38°S latitude through 6°S in New Guinea to 53°N latitude. Condensed data from these records is presented in tables and diagrams which compare overall totals, monthly totals, average daily totals and maximum midday exposure rates for the observing sites, all of which have special features concerning the incidence of skin cancer. A variety of other environmental influences have also been investigated experimentally, the summarized findings being presented herein. These include day to day variations, cloudiness, skylight, shade, altitude, reflection and transmission by common substances.The overall total environmental effective exposure does net change as rapidly with latitude as is commonly believed. Nevertheless measurements from clear skies are in reasonable agreement with the relative calculated values at various latitudes and times of year. Attention is drawn to the way in which other aspects of exposure vary with latitude in a way which can be expected to have independent influence on the rate of development of skin cancer. These include the intensity during the day, the duration of potential sunburning levels, the continuity of exposure, recovery processes and personal habits.From the accumulated physical and biological factors which appear to contribute significantly towards human sunburn and skin cancer, recommendations are made giving a framework on which preventive measures may be based. These are considered in terms of environment, clothing, external screening devices and the application of sunscreens to the skin. While each of these involves a great amount of practical consideration, they are dealt with briefly here with the exception of the topical use of sunscreens. Since this has attracted considerable public attention, the logic and the uncertainties of the use of sunscreens has been examined in detail, much of which is presented here as the prophylactic approach in which the public appears to show the greatest interest. The information presented in this work throws more light on the environment in which skin cancer develops. While it does not prove directly any new factor as a cause of skin cancer, it does give strong support for some known factors and helps to reduce the significance of others. Opportunity is provided for epidemiologists to re-examine old data with direct reference to exposure levels. The physical evaluation also suggests new types of population data which could be informative. It also suggests important factors which may be ascertained from skin cancer patients, making clinical records of greater value as a probe on carcinogenesis. A logical approach to protective measures, apart from generally reducing skin cancer, may be used to study carcinogenesis in highly susceptible persons. Suggestions are made for the long-term continuation of improved physical measurement to monitor the environment against possible increases in the ultraviolet end of the terrestrial solar spectrum.