The aim of the present case-control study was to ascertain whether, in adults, yearly repeated anti-influenza vaccinations (AIV) enhance protection against cutaneous melanoma (CM), as do repeated febrile infections. Ninety-nine new cases of histologically confirmed CM and 104 healthy controls (matched to cases for sex, age, and skin colour) selected from the general population were examined in order to ascertain their skin type, the number of nevi on both arms, and the intensity of freckles on the face and the arms; in these subjects, a structured questionnaire was used to obtain information on age, sex, education, social class, exposure and susceptibility to sunlight, history of febrile infectious diseases, and vaccinations. The odds ratio (OR) and the 95% confidence interval (CI) were estimated by commonly used methods and by fitting models of logistic regression. The risk of CM was reduced in subjects with a history of febrile (temperature above 38.5 degrees C) infections in the 5 years prior to CM surgery (cases) or interview (controls), but was increased in those with voluntary exposure to sunlight in tropical countries. By holding the above factors constant at logistic regression analysis, it was found that a history of repeated AIV (3-5 times in the last 5 years) halved the risk (OR: 0.43; CI: 0.19-1.00; p < 0.05). With the variable 'nevi on arms' included, the protective influence of repeated AIVs was observed in a similar magnitude. The inverse relationship found between melanoma and influenza vaccinations is unlikely to have depended on a bias, even if based on replies in a questionnaire, because neither the interviewers nor the interviewers were informed in advance of the working hypothesis.