The aim of this study was to try to identify those randomized controlled trials that compare sick leave with no sick leave or a different duration or degree of sick leave. A comprehensive, systematic, electronic search of Clinical Evidence, the Cochrane Library and PubMed, and a manual search of the Campbell Library and a journal supplement was carried out. Only two studies with randomization of sick leave were found. Patients, who were suspected of having whiplash injuries following a car accident, were randomly assigned to either 14 days on sick leave wearing a neck collar, or no neck collar or sick leave, and employees with musculoskeletal disorders were randomized to part-time sick leave and work adjustments, or full-time sick leave. Over 99% of the studies on sick leave were observational. The reason for the scarcity of randomized controlled trials is unknown. There is a definite, urgent need for randomized, controlled studies of the effects that sick leave has on health issues.