The aim of this systematic review and meta-analysis was to estimate the association between distinct types of child abuse--sexual (CSA), physical (CPA), and emotional (CEA)--and different eating disorders (EDs). Electronic databases were searched through January 2014. Studies reporting rates of CSA, CPA, and CEA in people with anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorder (BED), as compared with individuals without EDs, were included. Pooled analyses were based on odds ratios (ORs), with relevant 95% confidence intervals (CIs), weighting each study with inverse variance models with random effects. Risk of publication bias was estimated. Thirty-two of 1714 studies assessed for eligibility met the inclusion criteria, involving more than 14,000 individuals. The association between EDs and any child abuse showed a random-effects pooled OR of 3.21 (95% CI = 2.29-4.51, p < .001) with moderate heterogeneity (I2 = 57.2%, p = .005), whereas for CSA, this was 1.92 (95% CI = 1.13-3.28, p = .017), 2.73 (95% CI = 1.96-3.79, p < .001), and 2.31 (95% CI = 1.66-3.20, p < .001), for AN, BN, and BED, respectively. However, adjusting for publication bias, the estimate for CSA and AN was not significant (OR = 1.06, 95% CI = 0.59-1.88, p = .85). Although CPA was associated with AN, BN, and BED, CEA was associated just with BN and BED. BN and BED are associated with childhood abuse, whereas AN shows mixed results. Individuals with similar trauma should be monitored for early recognition of EDs. The protocol was registered in PROSPERO (an international prospective register of systematic reviews) with the reference number CRD42014007360.