Although errors can be a powerful impetus for learning, conventional pedagogy often emphasizes error-avoidance strategies that reward correct answers and disfavor mistakes. Error management training (EMT) takes an explicitly positive approach to errors, using them to create an active and self-directed learning environment. Using a surgical knot-tying model, we aimed to determine the efficacy of EMT among veterinary students with no prior surgical experience. We hypothesized that EMT would result in improved performance in unfamiliar scenarios (adaptive transfer) compared with an error-avoidance method. In this prospective double-blinded study, 42 students were equally divided between error avoidance training (EAT) and EMT groups. Performance in instrument- and hand-tied knots was evaluated for technique, time, number of attempts, and, when applicable, knot-leaking pressure. All participants demonstrated significant improvement between a pre-test and an analogous test 48 hours after training for all six outcomes (Wilcoxon matched pairs; two-tailed s ≤ .013). An adaptive transfer test found no significant differences between EMT and EAT at 48 hours (s ≥ .053). All participants demonstrated a significant performance decline in six of eight outcomes at 7 weeks post-training (s ≤ .021). This decline was not significant for four of six EMT outcomes yet significant for five of six EAT outcomes. These data suggest that students trained in both EMT and EAT experience comparable gains in short-term performance, including adaptive transfer. Compared with EAT, EMT may help attenuate performance decline after a sustained period of quiescence. Educators may consider actively incorporating EMT into veterinary curricula.