Mass screening for cutaneous melanoma is impracticable because of its low yield and high costs. The validity of self-selection of high-risk individuals through self-assessment of melanoma risk factors is not known. The aim of this study was to assess the validity of an unsupervised self-administered questionnaire for the self-assessment of melanoma risk. In a case-control study, melanoma cases (n = 202) and controls (n = 202) matched for age and gender filled in a questionnaire about melanoma risk factors. After filling in the questionnaire, all participants were interviewed and examined by a dermatologist in order to compare self-assessment with physician-assessment. The number of naevi, skin phototype and ultraviolet damage to the skin were identified as independent risk factors for melanoma on both self-assessment and physician-assessment. Receiver operating characteristics analysis showed no statistically significant difference between the accuracy of the self-assessment-based model and the model based on physician-assessment. While excluding 90% of the controls, the self-assessment-based high-risk group included 39% (95% confidence interval 31-48%) of the melanoma patients and the physician-assessment-based high-risk group included 42% (95% confidence interval 33-52%) of the melanoma patients. In conclusion, an unsupervised self-administered questionnaire is, to some extent, useful for the identification of individuals at high risk for melanoma. The moderate accuracy of self-assessment and physician-assessment to identify individuals at high risk for melanoma is a limitation for the practicability of targeted melanoma screening in general.