OBJECTIVES: To analyse the half-cycle correction and its effect on the final results of Markov models. METHODS: In our analysis we focus on the half-cycle correction, which is a method used to deal with the inaccuracy caused by inadequate cycle length in Markov models. The benefits of half-cycle correction has been widely published in the international literature. We measured the importance of half-cycle correction in the models submitted to the Hungarian HTA Office. We examined when it is adequate to use half-cycle correction and how big role should it have in the process of modelling. RESULTS: Our experience shows that only 11% of the submitted models incorporated half-cycle correction. In more than half of these cases the value of the incremental cost-effectiveness ratio (ICER) changed by less than 1% when half-cycle correction was used compared to the base-case scenario. We also found the possibility that in some cases the added benefit of half-cycle correction is not considerable. CONCLUSIONS: The necessity of using half-cycle correction is essential in models, when the cycle length is half year or longer and if the number of the cycles of the models is less than 200. In most cases the half-cycle correction results in only a little change in the cost-effectiveness ratio of the submitted models, therefore half-cycle correction should be executed carefully.