This article puts forward a new perspective of antitrust compliance based on a refined understanding of optimal deterrence. The traditional deterrence model implies that authorities ought to increase the level of fines and implement mechanisms (such as leniency policies) to increase the probability of detection in order to achieve an optimal level of antitrust deterrence and compliance. The traditional model nevertheless assumes that the corporate undertaking acts as a single, uniform entity instead of through multiple agents (ie directors and employees) with possibly diverse interests; the model thereby ignores the complexity of corporate mechanisms (of ex ante training and monitoring, and ex ante sanctions) that actually determine the level of antitrust compliance by a corporation. This article seeks to overcome this shortcoming by emphasising the constraint of corporate compliance costs, which must be traded-off against the diminishing benefits of added deterrence in determining the optimal level of antitrust penalty. Based on this refined understanding of optimal deterrence and compliance, this article criticises the UK judiciary’s simplistic deterrence-based reasoning in justifying the denial of a company’s recovery of its antitrust penalty by suing its delinquent agents for breach of duties.