Across numerous studies, individuals with substance use disorders (SUDs) differed from non-using controls regarding valuation of delayed gratification and feedback processing. However, it remains unclear whether the magnitude of the effect sizes is different across these two cognitive processes and how specific SUDs as well as demographic and clinical moderators influence these effects. In this study we thus performed multilevel linear mixed-effects meta-analyses and meta-regressions to examine the effects of SUDs on the Delay Discounting Task (DD) and on the Iowa Gambling Task (IGT). We found a moderate to large effect for SUD on both, the IGT and DD. While the effect on the DD was generalized to all substance classes, a smaller effect for cannabis-related disorder when compared to other SUDs was found with regard to the IGT. Early onset of substance use and psychiatric comorbidities were associated with stronger effects on the DD. Our findings suggest that feedback processing is more vulnerable to specific substance effects, while valuation of delayed gratification depends more on developmental and clinical factors.