The relation between use of sunscreens, different host factors and malignant melanoma was investigated in a population-based, matched case-control study of malignant melanoma in the South Swedish Health Care Region, which has the highest risk for melanoma in Sweden, between 1 July 1988 and 30 June 1990. In total, 400 melanoma patients and 640 healthy controls aged 15-75 years answered a comprehensive questionnaire regarding different epidemiologic variables, including questions on use of sunscreens and different constitutional factors. The use of sunscreens was not found to protect against developing malignant melanoma. Instead, an unexpected relation between the use of sunscreens and the risk of developing malignant melanoma was seen (odds ratio (OR) 1.8 for almost always vs never using sunscreens). A tentative dose-response relation was found. Virtually the same ORs were seen in both sexes. Furthermore, persons younger than 50 years had a higher OR than persons older than 50 years. When different melanoma presentation sites were considered, lesions of the trunk were associated with sunscreen use in females (adjusted OR = 3.7 for almost always vs never using sunscreens), while lesions of the extremity or head and neck were associated with sunscreen use in males (adjusted OR = 3.2 for almost always vs never using sunscreens). Raised naevi on the left arm and freckling were shown to be the major constitutional risk factors (OR = 3.9 for more than three naevi vs none and OR = 1.4, respectively). The results were essentially unaltered in a histopathologically re-examined material. Further investigations are needed in order to form a basis for melanoma prevention.