OBJECTIVE This study examined the predictors, extent and pattern of fatal antidepressant overdose (FAO) in all psychoactive drug-related deaths and compared these in drug abusers and non-drug abusers. METHODS Inquest data on 491 psychoactive drug-related cases in England and Wales were analysed. We examined predictors of FAO relative to other drug-related deaths in drug abusers and non-abusers using logistic regression, and compared the relative fatality of different antidepressant classes using the Proportionate Fatality Index (PFI). RESULTS The overall FAO prevalence ratio was 10% (male = 7.1%, female = 18.5%). FAO was twice as likely to occur among cases aged 45 years and over than in younger cases (MOR = 2.6, 95% CI = 1.4-4.9), and eight times more likely to occur in cases with no history of drug abuse (MOR = 8.2, 95% CI = 4.0-18.4). Ingestion of prescribed antidepressants is the only common predictor of FAO among drug abusers and non-abusers, and co-ingestion of a combination of antidepressants of two different classes is a potent risk for FAO in the cases studied. CONCLUSIONS The results highlight the need to develop appropriate therapeutic strategies for managing treatment-resistant depression.