In studies of skin cancer, participants are often classified into risk groups based on self-reported history of sun exposure or skin characteristics. We sought to determine the reliability of self-reported skin characteristics among participants of a study to evaluate markers for nonmelanoma skin cancer (NMSC). Multiple questionnaires and screening protocols were administered over a 3-month period to individuals from three risk groups: existing sun damage on forearms but no visible actinic keratoses (n = 91), visible actinic keratoses (n = 38), and history of resected squamous cell skin cancer in the last 12 months (n = 35). We assessed consistency of risk group assignment between telephone screen and study dermatologist assignment, self-reported sun sensitivity (telephone recruitment form versus participant completed profile), and self-reported history of NMSC skin lesions (telephone recruitment form versus health history). There was substantial agreement between probable risk group and final assignment (kappa = 0.76; 95% confidence interval, 0.65-0.85) and agreement did not differ by gender. Agreement for self-reported sun sensitivity was moderate (kappa weighted = 0.46; 95% confidence interval, 0.36-0.56) with higher agreement for women. For self-reported NMSC lesion history between two interviews, 24 days apart, kappa estimates ranged from 0.66 to 0.78 and were higher for women than men. Overall, there was evidence for substantial reproducibility related to risk group assignment and self-reported history of NMSC, with self-reported sun sensitivity being less reliable. In all comparisons, women had higher kappa values than men. These results suggest that self-reported measures of skin cancer risk are reasonably reliable for use in screening subjects into studies.