Social support is one of the most robust predictors of posttraumatic stress disorder (PTSD). Yet, little is known about factors that moderate the relationship between social support and PTSD symptom severity. This meta-analysis estimated the overall effect size of the relationship between self-reported social support and PTSD severity and tested meaningful demographic, social support, and trauma characteristics that may moderate this association using both cross-sectional and longitudinal effect sizes. A comprehensive search identified 139 studies with 145 independent cross-sectional effect sizes representing 62,803 individuals and 37 studies with 38 independent longitudinal effect sizes representing 25,792 individuals. Study samples had to comprise trauma-exposed, nonclinical adult populations to be included in the analysis. Cross-sectional and longitudinal analyses revealed a near medium overall effect size (rcross = -.27; 95% CI [-.30, -.24]; rlong = -.25; 95% CI [-.28, -.21]) with a high degree of heterogeneity (cross-sectional I2 = 91.6, longitudinal I2 = 86.5). Both cross-sectional and longitudinal moderator analyses revealed that study samples exposed to natural disasters had a weaker effect size than samples exposed to other trauma types (e.g., combat, interpersonal violence), studies measuring negative social reactions had a larger effect size than studies assessing other types of social support, and veteran samples revealed larger effect sizes than civilian samples. Several other methodological and substantive moderators emerged that revealed a complex relationship between social support and PTSD severity. These findings have important clinical implications for the types of social support interventions that could mitigate PTSD severity. (PsycInfo Database Record (c) 2021 APA, all rights reserved).