Stressors like test anxiety are known to decrease memory retrieval, whereas retrieval practice is the phenomenon that actively recalling information from memory enhances memory. Evidence suggests retrieval practice can protect memory against the negative effects of stress on memory (Agarwal et al., 2014; Smith & Thomas, 2018), however the findings are mixed (Yang et al., 2020). Understanding whether retrieval practice maintains memory performance under stress could transform memory resilience and lead to new cognitive interventions. This raises the need for a meta-analytic summary of the literature to understand the effects of retrieval practice on memory in relation to stressors. In this registered report, we conducted a meta-analysis (k = 10, n = 908) on the impact of stress following learning with retrieval practice from 2006–2024 using databases and other information sources. We found a positive effect of retrieval practice in stress conditions, Hedges’ g = 0.45, 95% CI [0.17, 0.72] with a random-effects two-level model, suggesting retrieval practice is more beneficial than restudy during stress. Significant heterogeneity was observed along with a moderate risk of bias, yet little evidence for publication bias. We did not find significant moderators in this sample. Future studies should be conducted with sufficiently powered samples and using other types of control, or non-retrieval practice, strategies to better understand the protective effects of retrieval practice on memory following stress. We registered our meta-analysis, with datafile, code and supplementary here: https://osf.io/jwx4f/.