|Abstract:||Behavioral priming traces its roots back to 1890, but recently has been criticized for failure to replicate. In response, researchers argued the need for more meta-analyses (Bargh, 2012). Accordingly, the following meta-analysis sought to estimate its mean effect size and to quantify publication bias as well as the effects of moderators like goal expectancy (ease), filler tasks, goal value, and the directness of the association between prime and outcome. We estimated a mean of d = 0.4220 across 682 eligible effect sizes. Despite evidence of some publication bias, the trim and fill procedure recalculated the mean effect to be d = 0.2661, indicating that behavioral priming is likely a true effect, if perhaps overestimated in the literature. Our findings regarding goal expectancy, filler tasks, and goal value are somewhat mixed, but we consistently found that indirect primes, which require more interpretation to produce the outcome, had larger effect sizes than more direct primes.