Abstract: If there are gross errors or cycle slips in the Global Navigation Satellite System (GNSS) comprehensive corrections in observation space representation, it will inevitably affect precise point positioning (PPP) augmented performance. Aiming at the possible anomalies in the comprehensive corrections, a quality control algorithm for the comprehensive corrections was studied and an algorithm for identifying and controlling outliers was proposed. According to the characteristics of the comprehensive corrections, the median absolute deviation was used to identify and locate the outliers for the combined value of comprehensive corrections after inter-frequency difference and second-order inter-epoch. The satellites with outliers used were control by the means of ambiguity reinitialization, weight reduction or elimination to reduce the impact of outliers on the results. The method was validated with a reference network composed of some Hong Kong continuously operating reference stations (CORS) with an average side length of 26 km and zero baseline data from Curtin University. The results showed that the method can effectively detect most of the large anomalies more than 1 cycle and some anomalies within 1 cycle n the value of the differential combination, and effectively control the influence of some outliers on the positioning results.