Background: The present study meta-analytically reviewed the effects of exercise on four transdiagnostic treatment targets: anxiety sensitivity (AS), distress tolerance (DT), stress reactivity (SR), and general self-efficacy (GSE). Methods: We conducted systematic searches of peer-reviewed studies in bibliographical databases (Cochrane Library, psychINFO, PubMed) before April 1, 2018. Only randomized controlled trials (RCT) evaluating the effect of exercise on AS, DT, SR, or GSE using at least one validated outcome instrument in a sample of adolescents (≥13 years old) or adults were selected. We employed a meta-analysis of effects using random-effects pooling modeling for each treatment target. Results: The systematic search yielded 28 RCTs meeting eligibility criteria. Exercise interventions had a large effect on reducing AS (six studies, Hedges's g = 0.72, p = .001), a medium effect on increasing GSE (eight studies, Hedges's g = 0.59, p < .001), and a small effect on reducing SR (ten studies, Hedges's g = 0.32, p < .001). Evidence from four studies suggested that exercise interventions had a small but non-significant effect on increasing DT (Hedges's g = 0.21, p = .26). Conclusions: This meta-analysis provides preliminary evidence exercise can engage certain transdiagnostic targets. Further research is required to optimize exercise intervention parameters to achieve the strongest effects on these important mechanistic variables.