In this paper, a methodology is presented for machine tool error correction which uses reference parts as metrological standards. These reference parts have the same geometric shape and weight as. real workpieces. Consequently, the proposed methodology minimizes the uncertainty which arises from geometric interpolation, but still yields satisfactory results for multi-axis machine error characterization. A generalized error model is derived 10 describe the relative error motion between the 1001 and the workpiece. Iterative search algorithms arc presented which increase the effectiveness of the error charactcrization process. Three case studies are described which demonstrate the application of the methodology for foul' sample reference parts on three different machines. The results from these studies show that the error correction method can be used to characterize the positioning errors with estimation accuracy approaching the resolution of the machines.