Research suggests that stimuli that prime social concepts can fundamentally alter people’s behavior. However, most researchers who conduct priming studies fail to explicitly report double-blind procedures. Because experimenter expectations may influence participant behavior, we asked whether a short pre-experiment interaction between participants and experimenters would contribute to priming effects when experimenters were not blind to participant condition. An initial double-blind experiment failed to demonstrate the expected effects of a social prime on executive cognition. To determine whether double-blind procedures caused this result, we independently manipulated participants’ exposure to a prime and experimenters’ belief about which prime participants received. Across four experiments, we found that experimenter belief, rather than prime condition, altered participant behavior. Experimenter belief also altered participants’ perceptions of their experimenter, suggesting that differences in experimenter behavior across conditions caused the effect. Findings reinforce double-blind designs as experimental best practice and suggest that people’s prior beliefs have important consequences for shaping behavior with an interaction partner.