Substance use disorders are characterized by marked changes in reward and error processing. The primary objective of this meta‐analysis was to estimate effect sizes for the reward positivity (RewP) and error‐related negativity (ERN), two event‐related potential indicators of outcome monitoring, in substance users compared to controls. The secondary objective was to test for moderation by demographic, substance type, and EEG experiment parameters. Final PubMed searches were performed in August 2023. Inclusion criteria were substance use disorder/dependence or validated self‐report of substance misuse, RewP/ERN means available, healthy control comparison group, non‐acute drug study, peer‐reviewed journal, English language, and human participants. Selection bias was tested through modified Egger's regression and exploratory 3‐parameter selection model tests. The RewP results (19 studies, 1641 participants) did not support an overall effect (Hedges' g = 0.07, 95% CI [−0.44, 0.58], p = .777) and nor effect of any moderators. The ERN results (20 studies, 1022 participants) indicated no significant overall effect (g = 0.41, 95%CI [−0.05, 0.88]). Subgroup analyses indicated that cocaine users had a blunted ERN compared to controls (g = 1.12, 95%CI [0.77, 1.47]). There was limited evidence for publication/small study bias. Although the results indicate a potential dissociation between substance types, this meta‐analysis revealed the need for additional research on the RewP/ERN in substance using populations and for better designed experiments that adequately address research questions.