We examine micro-level data on medical incidents recorded by a major Belgian professional liability insurer to identify the predictors of medical malpractice victim compensation. The data allow us to track each instance of suspect medical malpractice from the moment of insurer's knowledge about the incident to file closure. We are, therefore, able to investigate the determinants of both the incidence and amount of indemnity payment while addressing the associated sample selection concerns. Conditional on some indemnity having been paid, provider specialty risk predicts the indemnity payment amount, but only via the effect of sustained injury type. We find little evidence of vertical or horizontal inequities in indemnity payment. Our results highlight previously overlooked features of the incident resolution process as quantitatively important predictors of victim compensation.