With the development of industries and the advanced manufacturing technology, the performance of the machine tools plays an important role for the product quality. Machining performance of five-axis machining centers, MC, in the manufacturing industry has been a hot research area, where evaluation methods based on cutting of a test piece are common, but it have proved hard to determine the relationship between test piece deviations and machine tool sensor data during manufacturing. An "S" test specimen has the characteristics of open and close angle, variable curvature, thin wall, and low stiffness. The quality of a machined "S" specimen is used to test the characteristics of five-axis MCs. Inaccurate manufacturing results in unqualified parts and the loss of business. Normally geometrical errors are hard to adjust and compensate for in the process. In order to evaluate, and map the relationship between the characteristics of geometrical work-piece deviations from a MC to in-line sensor data, a monitoring system was established in this study. The LABVIEW monitoring evaluation system is developed based on "running performance tracing". By using a coordinate measuring machine, CMM, and a Scanning Electron Microscope to measure the actual geometry and surface finish of the machined S part and compare it to the ideal geometry, a database and a running performance evaluation model for 5-axis MC based on the "S" specimen was developed. Finally, the effectiveness of the method of evaluation is verified by cutting experiments using a MAZAK five-axis MC. The Achievements reported here are helpful for solving the lack of systems to evaluate the running performance of five-axis CNC machine tools for high-end manufacturing industry in China, which has important application prospect and high economic value.