The weapon identification task (WIT; Payne, 2001) is a prominent implicit measure of racial stereotyping. In this experimental sequential priming paradigm, participants classify objects (weapons vs. tools) following brief presentations of Black or White male faces. The so-called stereotype effect in the WIT reflects faster and/or more accurate responses to Black–gun and White–tool pairings than to Black–tool and White–gun pairings. Previous meta-analytic summaries (Rivers, 2017; Kidder et al., 2018) have corroborated this effect but have certain constraints, such as unexplained heterogeneity, limited search coverage, and the absence of moderator analyses. To provide a more comprehensive and systematic assessment, we conducted a meta-analysis of 35 WIT studies conceptually aligned with Payne’s (2001) original paradigm. In addition to estimating overall stereotype effects in mean response times and error rates, we examined potential sources of between-study heterogeneity, focusing on response-deadline strictness, participant gender, and publication year. Results revealed significant small-to-medium stereotype effects for mean response times (g_z = 0.39, 95% CI [0.29–0.49]) and error rates (g_z = 0.42, 95% CI [0.33–0.51]). Deadline strictness moderated the stereotype effect for error rates only, with stronger effects observed under stricter deadlines. Furthermore, the stereotype effect increased with the proportion of female participants across both dependent measures. Finally, the stereotype effect declined across publication years, though only for mean response times. These findings offer transparent, reproducible, and up-to-date evidence for stereotype effects in the WIT while highlighting design- and sample-related moderators that may account for previous heterogeneity.