Background Acute stress disorder (ASD) was proposed to encapsulate traumatic stress reactions within the first few months of exposure to trauma. The present systematic review and meta-analysis aimed to estimate the prevalence of ASD in children and adolescents, and the extent to which assessment, demographic and trauma variables moderate this. Method Searches of EMBASE, MEDLINE (PubMed), PsycINFO, PsycARTICLES and PILOTS were conducted to identify studies published between 1st January 1994 and 1st January 2018. Seventeen studies were identified as meeting inclusion criteria (N=2918 participants). Results The pooled prevalence estimate for ASD was 16.5% (95% CI 10.6–23.4%), with considerable heterogeneity between studies (Q=261.12, p < .001, I2=95.3%). Risk of bias was unrelated to prevalence estimates. Studies that used a clinical interview (k=8) yielded a higher estimate (24.0%, 95% CI 13.8–36.0%) than those that used a questionnaire which adhered to the diagnostic algorithm for DSM-IV ASD (k=6; 6.8%, 95% CI 3.6–10.9%). Studies comprising older participants yielded greater prevalence estimates. Prevalence was significantly greater in studies where the majority of participants had been exposed to interpersonal trauma (27.9%, 95% CI 15.1–42.8%; k=5) compared to non-interpersonal trauma (12.8%, 95% CI 7.2–19.7%; k=12). Conclusions This review suggests that a significant minority of trauma-exposed children and adolescents meet criteria for ASD (in particular youth exposed to interpersonal trauma), but the findings are limited by a large degree of heterogeneity. DSM-IV ASD-specific self-report questionnaire measures may be too insensitive for identifying youth with this disorder.