OBJECTIVES: In economic models that use Markov-type processes, it is generally recommended that a ‘half-cycle correction’ be built into the analysis, to account for the fact that events can occur at any point during the cycle.  This study explores the implications of the half-cycle correction, and highlights a number of flaws in the approach. METHODS: A brief review of health technology assessment models was undertaken to determine the use of half-cycle corrections.  The study aimed to explore the theoretical, practical and mathematical implications of the half-cycle correction.  A simple Markov model was built to illustrate the impact of the half-cycle correction, and to demonstrate how a more accurate correction factor can be applied to models. RESULTS: Half-cycle corrections appear to be used routinely in Markov models.  In nearly all cases, the so-called “correction” is applied without due consideration of the implications.  Two major flaws were identified with the approach.  The first, mathematical, flaw is that the half-cycle correction approach assumes that all events occur at the mid-point of each cycle.  It can be demonstrated that, for one-directional events (such as death), events will be more likely to occur in the first half of the cycle since more patients will be exposed to the event at the start of the cycle, and the number of patients ‘at risk’ falls throughout the cycle.  The second flaw is that, for many events, the implications of the event may not actually become apparent until the next cycle.  For instance, in oncology, the increased costs associated with disease progression will not occur until progression is confirmed, which may only happen at regular routine follow-up visits. CONCLUSIONS: Half-cycle corrections are frequently applied inappropriately in modelling.  This study has produced two key recommendations to generate more accurate outcomes and to avoid biases in decision analytic models.