530 patients with localised cutaneous melanoma consecutively observed between 1980 and 1991 at a hospital for skin diseases in Rome, Italy, were studied. Crude survival proportions were calculated with the method of Kaplan and Meier. Cox proportional hazards regression analysis was used to estimate the effect of prognostic factors on death rates. Females and younger patients had better 5- and 10-year survival rates, while increasing tumour thickness was associated with a decrease in survival time. In the multivariate analysis, an independent association with survival was found for tumour thickness, presence of ulceration, age, sex and cross-sectional profile of neoplasia. Our study confirms that females and young patients with thin melanomas have a better prognosis, while the importance of cross-sectional profile needs further study.