Abstract: To address the issues of indicator redundancy and the subjective nature of weight determination in the accuracy evaluation of high-end CNC machine tools, this study proposes a combined weighting and key indicator screening method that integrates the analytic hierarchy process (AHP) with information entropy. Firstly, a multi-level indicator library covering geometric, positional, machining, and motion accuracies was systematically compiled. Secondly, a combined weighting approach was developed by integrating AHP weights derived from expert experience with entropy weighting based on empirical measurement data. By applying the Pareto principle, key indicators accounting for approximately 80% of the cumulative weight were selected based on their cumulative contribution, thereby forming a streamlined evaluation system that highlights critical aspects. When applied to evaluate five machine tools of different types, the comprehensive scores and the distribution of indicator strengths and weaknesses aligned with engineering expectations, validating the system’s effectiveness and practicality. This method provides a practical theoretical tool for the systematic evaluation and quality classification of machine tool accuracy.