The minimal cut set approach is extensively used in probabilistic safety assessments. Truncation neglecting low-probability cut sets is usually applied to determine the minimal cut sets of a manageable size. The errors in the calculated top event probability due to the truncation are not usually known. A truncation error evaluation method, which is based on the Monte Carlo simulation technique and the characteristics of coherent reliability systems, is proposed in this paper. This method can be easily applied to the most problems in probabilistic safety assessments and does not require much computation time. The applicability and accuracy of the proposed method are also described.