In GPS and GLONASS satellite-based positioning, the correct determination of the integer carrier phase ambiguities is the key to achieve precise positioning results. The process of determining the ambiguities, called ambiguity resolution, is usually separated into three steps: (a) estimating the float or real-valued ambiguities, (b) searching the best integer ambiguity set, (c) validating the best ambiguity set. Whilst the theory and methodology for ambiguity estimation and search are well documented, the ambiguity validation issue is still under investigation. Existing ambiguity validation procedures are based on various statistical assumptions and therefore, may have different performances in practical applications. In this paper, the major ambiguity validation procedures are analysed and their performances in both GPS and GLONASS ambiguity resolution are numerically compared. It is concluded that both the probability of estimating ambiguities correctly and confidence levels of the ambiguity discrimination tests are both important indicators of the reliability of ambiguity resolution.