The effect of age, cohort, and year of death on the mortality of cutaneous malignant melanoma is determined by the use of a Poisson log-linear model. During the period of the study, mortality due to this tumor increased exponentially (an annual rise of 11% in both sexes). The model attributes this rise to a cohort effect. The relative risk for the 1952 cohort as compared to the 1892 cohort is 530 for men and 280 for women. In the younger generations, no signs of leveling off are to be seen. The sharp increase in mortality due to malignant melanoma of the skin has also been witnessed in other countries and suggests a real increase in incidence. One explanation for this epidemic phenomenon lies in the progressive rise in exposure to UV radiation.