Solar exposure, vitamin D, and their possible beneficial effect on cancer risk and cancer prognosis are a topic for research. Despite the distinct nature of sunlight, it has proved difficult to assess the exposure quantitatively in epidemiological studies. Skin cancers, latitude, and sunny climate have been used as proxy indicators of solar exposure above a reference level. The interpretation of such data may still be hampered by incomplete cancer registration, difference in protection against sunbeams, selection mechanisms, and absence of information on potential confounders. A recently published paper -- on second primary cancer following the diagnosis of a skin cancer -- is discussed to illustrate the difficulties. Further epidemiological studies of potentially protective effects from carcinogenic ultraviolet rays should include individual information on solar exposure and vitamin D levels, as well as on other recognised and relevant risk factors.