Pesticides help produce food for humanity's growing population, yet they have negative impacts on the environment. Limiting these impacts, while maintaining food supply, is a crucial challenge for modern agriculture. Mitigation measures are actions taken by pesticide users, which modify the risk of the application to nontarget organisms, such as bees. Through these, the impacts of pesticides can be reduced, with minimal impacts on the efficacy of the pesticide. Here we collate the scientific evidence behind mitigation measures designed to reduce pesticide impacts on bees using a systematic review methodology. We included all publications which tested the effects of any pesticide mitigation measure (using a very loose definition) on bees, at any scale (from individual through to population level), so long as they presented evidence on the efficacy of the measure. We found 34 publications with direct evidence on the topic, covering a range of available mitigation measures. No currently used mitigation measures were thoroughly tested, and some entirely lacked empirical support, showing a weak evidence base for current recommendations and policy. We found mitigation measure research predominantly focuses on managed bees, potentially failing to protect wild bees. We also found that label-recommended mitigation measures, which are the mitigation measures most often applied, specifically are seldom tested empirically. Ultimately, we recommend that more, and stronger, scientific evidence is required to justify existing mitigation measures to help reduce the impacts of pesticides on bees while maintaining crop protection.