Immune function plays a prominent role in the defence against cutaneous malignant melanoma and the increased risk of melanoma development during immunosuppression. Since the immune system is challenged beyond its routine activity by an infection, the effect of previous infectious diseases on the risk of melanoma may also be crucial. In a European Organization for Research and Treatment of Cancer (EORTC) case-control study performed in six European countries and Israel, we compared the history of severe infections in 603 melanoma patients with that in 627 population controls. We calculated adjusted odds ratios (ORs) to estimate the effect of infectious diseases on melanoma risk. The ORs for melanoma risk were below 1 for nearly all types of infections (except two) if body temperature was not taken into consideration, and for all infections with a body temperature above 38.5 degrees C. In the latter category significantly lowered ORs were found for pulmonary tuberculosis (0.16; 95% confidence interval [CI] 0.01-0.98), Staphylococcus aureus infections (0.54; 95% CI 0.31-0.94), sepsis (0.23; 95% CI 0.06-0.70), influenza and related infections (0.65; 95% CI 0.48-0.86) and pneumonia (0.45; 95% CI 0.27-0.73). Analysis of the cumulative influence revealed a consistent pattern of results pointing to a reduction in melanoma risk with increasing numbers of recorded infections and fever height. This apparent dose-response relationship suggests a causal association. Speculations on the underlying mechanism include a Shwartzman-like phenomenon when melanoma formation precedes the infection and/or an infection-related Th1-cell activation preventing the establishment of the tumour.