The ethical principle of ‘equipoise’, introduced in 1974, represents the most widely influential justification for the enrollment of patients into randomized clinical trials. However, definitions of equipoise vary, and its terms are not universally accepted. In this paper, we suggest a new way of approaching the ethics of clinical trial enrollment, which we call fallibility. The principle of fallibility argues that all physician opinions are sufficiently uncertain to warrant investigation, and that the ethical justification for any trial becomes a question of its epistemic validity, by which we mean the strength of its hypotheses and methods. The principle of fallibility can be translated into practice through the virtues of humility, skepticism and caring. While we cite recent examples from stroke medicine to demonstrate the limitations of equipoise, we propose that fallibility may offer a more general means of addressing the controversies that arise surrounding randomized controlled trials in many disciplines of medicine.