Many experimental researches have been conducted until recent years to see the effect of mastery learning approach on students’ cognitive behavior and affective characteristics. But the question is how much evidence is there in the existing research results provides scientific conclusions by combining existing experimental results. By treating different experiments of mastery learning as research replications, the experimental results can be combined using a meta-analysis technique. This paper shows how a quantitative research synthesis can effectively be used to combine statistical evidences of researches conducted separately and independently. The effect of mastery learning on affective characteristics of students was selected for this research synthesis. The mastery learning approach to be investigated in this research synthesis is Bloom type of mastery learning strategy. Using 26 independent comparisons, the results of study show that: a) the effect sizes of mastery learning on affective characteristics of students are heterogeneous across studies, b) the source of study, either from dissertation or journal article, does not explain the variability among the effect sizes, c) mastery learning programs using a ≥ 75% mastery criterion seem to have positive affective impact on the students, while those using < 75 % mastery criterion have no impact on the affective characteristics of students, d) the mean effect size shows a decreasing trend as the level of education increases, e) the mean effect size is highly positive for mathematics class and low positive effect for science and social studies, and f) short treatment duration has a much larger positive effect size than the long term treatment duration.   DOI: 10.15408/tazkiya.v20i2.9176