Résumé : Experiments in rodents and humans have shown the ability of sunscreens to prevent UVB-induced skin cancers. Sunscreen use has thus become part of the panoply for melanoma prevention. However, results from epidemiological investigations indicate that sunscreen use could be a risk rather than a protective factor for melanoma. This article discusses possible explanations for discrepancies between experimental and observational results, suggesting that experiments did not reproduce usual human behaviours towards the sunlight. Also, marketing of sunscreens fosters unconsidered exposure to sunlight through associating sunscreen use to the acquisition of a 'safe suntan', or to the possibility to let children go naked into the sunlight without incurring sunburns. It has therefore been hypothesized that, because it prevents sunburn, sunscreen use allows prolonged exposure to sunlight. If this hypothesis holds, modern sunscreens with high protection factor could also represent a threat if their use is motivated by the desire of unrestricted exposure to sunlight. To explore this hypothesis, the design of a retrospective cohort study is presented, including children 5-7 years old, with the total body naevi count as principal endpoint. This study will have the opportunity to examine the qualities of modern sunscreens among young children.