The aim of this study is to identify-by merging the results of a large number of studies conducted in related literature review-at which level "writing as a process" approach affects students' writing success. Additionally, this paper investigates whether the writing success level differentiates depending on certain study characteristic. Metaanalysis has been preferred as research method in this study and among the studies which are associated with process-based writing practice, only the results of 21 experimental/quasi experimental studies that are conducted in Turkey have been synthesized which meet the inclusion criteria. In this regard, effect size measures (Hedge g) of studies included into meta-analysis, heterogeneity, publication bias tests and intervening variable analysis have been done with the use of Comprehensive Meta Analysis v2.0 (CMA) software. In the results of study, Hedge g=0,983 has been determined as the corrected effect size related to effect of process-based writing approach on students' writing success levels. This value indicates that process-based writing approaches affect students' writing success levels to a large extent. The results of this study reveal that-in spite of teaching levels-process based writing model, duration of practice, text type used in practice and publication type in which practice has been reported do not make a statistically significant difference in effect sizes.