Controversy exists surrounding the role of childhood abuse in obesity development. This is a meta-analysis of observational studies on the role of childhood abuse in adult obesity. Systematic searches of PubMed, PsycInfo, Medline and CINAHL resulted in 23 cohort studies (4 prospective, 19 retrospective) with n = 112,708 participants, containing four abuse types (physical, emotional, sexual, general). Four studies reported dose–response effects. A random effects model was used to quantify effect sizes, meta-regression/subgroup analysis for identifying potential moderating variables and Egger's test for publication bias. Adults who reported childhood abuse were significantly more likely to be obese (odds ratio [OR]: 1.34, 95% confidence interval [CI]: 1.24–1.45, P < 0.001). All four types of abuse were significantly associated with adult obesity: physical (OR: 1.28, 95% CI: 1.13–1.46), emotional (OR: 1.36, 95% CI: 1.08–1.71), sexual (OR: 1.31, 95% CI: 1.13–1.53) and general abuse (OR: 1.45, 95% CI: 1.25–1.69). Severe abuse (OR: 1.50, 95% CI: 1.27–1.77) was significantly more associated with adult obesity (P = 0.043) compared with light/moderate abuse (OR: 1.13, 95% CI: 0.91–1.41). We found no significant effects of study design (prospective vs. retrospective, P = 0.07), age (P = 0.96) or gender (P = 0.92). Publication bias was evident (Egger's test P = 0.007), but effect sizes remained statistically significant in sensitivity analyses. Childhood abuse was clearly associated with being obese as an adult, including a positive dose–response association. This suggests that adverse life experiences during childhood plays a major role in obesity development, potentially by inducing mental and emotional perturbations, maladaptive coping responses, stress, inflammation and metabolic disturbances.