The statistical tests of the post fit residuals are applied to the regional GPS reduction. The strategies of fixed precise orbits and precise point positioning are compared. It is shown that in the former case, conceivable signals, still remain in the post fit redsiduals while in the latter case they are mainly noises. Based on the differences between our solutions with ITRF96 and the repeatabilities of the solutions, it is concluded that the stragegy with reasonable post fit residual statistics is better than others, Statistical tests of the post fit residuals present another evidence that besides accuracy and computation saving, PPP is more preferable than fixed precise orbits in the regional GPS analysis.
