Considering the complex tracking measurement system and high development cost of active laser tracker, a passive laser tracking method is proposed to measure the spatial coordinates of target points. Firstly, the error analysis model of the passive laser tracking measurement system was established according to HTM principle, and the detection method of relevant error parameters was proposed. The measurement error of the passive laser tracking measurement system was reduced from 550μm to 150.8μm after error compensation. In order to further improve the measurement accuracy, residual errors after error compensation were collected as sample data, and BP neural network was used to train the sample data. After the residual error of the passive laser tracking system was compensated by the trained BP neural network model, the measurement error of spatial coordinate measurement was reduced from 150.8μm to 51.6μm. Compared with HTM error compensation method, the compensation effect of HTM+BP neural network model is improved by 65.8%.