In this research, a meta-analysis study was conducted in order to determine the effects of constructivist learning approach on students' academic achievement. Master's thesis, doctoral dissertation and articles in national and international databases, which are realized between the years of 2003-2014, appropriate to the problem and which can be included in a work of meta-analysis with important statistical data, have been studied by scanning in Turkish and English. At the end of the literature review, a total of 53 studies about effects of constructivist learning approach on student's academic achievement have been included in the meta-analysis study. Meta-analysis study determined that the constructivist learning approach, compared to traditional teaching methods, has positive effects on the student's academic achievement. After using a random effects model with a 0.910 and 1.402 interval of the confidence, the overall effect of constructivist learning approach in relation to academic achievement of students is found 1.156 (95% CI, SE = 0.125). 50 of the 53 studies included in the study have positive results, though only 3 of them show negative effect. As a result of moderator analyzes, we see that the highest effect values are observed in the master's thesis, in the teaching of science and at the college level.