An off-axis paraboloid with an off-axis angle of 45° and a diameter of 1 inch was used as the sample to compare the difference between the traditional null interferometric measurement results and the non-contact 3D aspheric optical surface shape measurement system Luphoscan. Zemax was used to simulate the influence of the adjustment errors of the six dimensions of the off-axis paraboloid on the measurement results. The expression of off-axis error in the two completely different measurement processes was analyzed, which solved the problem of the availability of Luphoscan measurement results. According to the simulation results, the null test interference optical path structure based on the dynamic interferometer was set up, and the six-dimensional adjustment standard process of the off-axis paraboloid was established, which could quickly eliminate the adjustment error and off-axis quantity error, and the repeatability test results reach the accuracy of pv=0.135λ.