Non‐consumptive effects (NCEs) – changes in prey behavior or physiology in response to predator threat – are common and can be as strong as consumptive effects. However, our knowledge of NCEs in arthropod systems is lacking. Factors related to study organism and environment have the potential to influence the occurrence and magnitude of NCEs in arthropod systems. While factors such as coevolutionary history of natural enemies and their prey, predator cue, predator or prey feeding mode, and refuge availability have been theoretically and empirically examined, no trends have been proposed for arthropods. We compiled 62 studies, yielding 128 predator–prey interactions, which explicitly examined NCEs in experiments where arthropods were identified to species, using a previously published database of papers from 1990 to 2005 and a new database of papers published from 2006 to 2015. Using these data, we conducted a meta‐analysis to explore the influence of organismal and environmental characteristics on the magnitude of predator NCEs. Our analysis addressed the following three questions. 1) Does predator–prey coevolution give rise to stronger NCEs than when predator and prey species did not coevolve? 2) What influence does habitat type and refuge availability have on NCEs? 3) How do predator characteristics (cue type, hunting mode and life stage) and prey characteristics (mobility, life stage, specialization, gregariousness and feeding mode) influence NCEs? We found that while NCEs were similar across most measured characteristics, NCEs on prey activity were significantly stronger when predator and prey shared an evolutionary history. Our results support growing evidence that NCEs have a negative effect on prey traits and that behavioral NCEs are stronger than physiological ones. Additional studies are needed to be confident in any emerging patterns, therefore we identify key gaps in the literature on NCEs in arthropod systems and discuss ideas for moving forward.