<p>The effect of medical malpractice liability on physicians’ behavior remains poorly understood as does its interaction with payment systems. In this paper, we examine whether the effectiveness of a standard, optimally calibrated, mixed payment system in inducing an efficient (socially optimal) level of service provision is influenced by the presence of malpractice liability. Using a controlled laboratory setting, we investigate the role of medical malpractice liability on physicians’ service provision under capitation, fee-for-service, and mixed payment systems. We consider four between-subject treatments, each referring to a pure payment system (either fee-for-service or capitation) and a mixed one. Each treatment includes four within-subject conditions that differ in payment systems (pure vs. mixed), the order in which the pure and mixed systems are played and the presence/absence of medical liability (no liability vs. liability). We find that while the mixed payment system mitigates the under- and over-provision induced by capitation and fee-for-service, respectively, the introduction of liability pressure leads to a significant deviation from the optimal level of treatment that is not corrected by a standard mixed system. Our empirical findings confirm the existence of defensive medical practices and indicate that malpractice liability has substantial policy relevance for the optimal design of physician payment systems. Moreover, under capitation, liability—when combined with physician altruism—may align incentives with socially efficient care, generating outcomes comparable to those of an optimally calibrated mixed payment scheme.</p>