Effective monitoring is necessary to provide robust detection of bee declines. In the United States and worldwide, bowl traps have been increasingly used to monitor native bees and purportedly detect declines. However, bowl traps have a suite of flaws that make them poorly equipped to monitor bees. We outline the drawbacks of bowl traps, as well as other passive sampling methods. We emphasize that current methods do not monitor changes in bee abundance. We then propose future approaches to improve bee monitoring efforts, which include improving our understanding of the efficacy and drawbacks of current methods, novel molecular methods, nest censusing, mark-recapture, sampling of focal plant taxa, and detection of range contractions. Overall, we hope to highlight deficiencies of the current state of bee monitoring, with an aim to stimulate research into the efficacy of existing methods and promote novel methods that provide meaningful data that can detect declines without squandering limited resources.