One of the most frequently studied phenomena in educational researches is how educational stakeholders (teachers, administrators, supervisors, academicians) experience burnout. Maslach and Jackson (1981), the developers of mostly used inventory in the literature, suggested that burnout is a multidimensional construct. It consists of three dimensions: emotional exhaustion, depersonalization, and (reduced) personal accomplishment. The first dimension emotional exhaustion (EE) results from feelings of intensity and frustration as a result of individuals" fears. When people experience burnout emotionally, they will not be able to enhance previous levels of their job performance. The second dimension is depersonalization (DP) and it occurs when people hold off themselves from their work environment by creating depersonalizing perceptions of tasks, and/or coworkers (Kahn, Schneider, Jenkins-Henkelman and Moyle, 2006; Schaufeli and Enzmann, 1998). The third dimension of burnout is reduced personal accomplishment (PA), defined as self-evaluative feelings of inadequacy and lack of achievement at work environments (Maslach and Leiter, 2008). Thus, high levels of emotional exhaustion and depersonalization together with low levels of personal accomplishment are symptoms of burnout. In this study, a meta-analytic method was used to synthesize gender differences in burnout at educational settings of Turkey. The magnitude and direction of the difference between burnout and gender was examined. Practitioners will use the findings of this study to help identify individuals who are at greater risk of burnout. If possible, early prediction would allow schools to screen out individuals who are predisposed to burnout and implement preventive interventions more effectively (Maslach and Leiter, 2008) in an effort to decrease their psychological and physiological disturbances and increase their performance. A synthesis result of this study may direct the future researches on the subject in terms of gender. In order to identify all potential studies for inclusion in the quantitative synthesis, a comprehensive systematic search strategy was used. As the databases, ULAKBIM (Turkish Academic Network and Information Center) Social Sciences Database, YOK (Higher Education Council) Doctoral and Master Theses Database, Inonu University Library Central Search, several databases by Ebscohost such as ERIC and HIC, PsycArticles, Emerald, Jstor, Web of Science and Google Scholar were searched. The stems of following identifiers or keywords in the title or abstract were used in the separate or combined searches: (In Turkish) tükenmi?*, duyarsızla?ma, ki?isel ba?arı*, eğitim*, cinsiyet* ölçek, (In English) burnout, exhaustion, depersonalization, personal achievement, Maslach Burnout Inventory. The search included studies available in the academic literature from 2004 until 2014. From this search, 3489 studies were retained for further inspection. Compatible with the inclusion criteria only 75 of these studies were included in the analyses. In order to be included in the analysis, studies had to meet five specific criteria. 1-The study examined the association between burnout and gender and it had to report data with gender differences in all three dimensions of burnout using a version of the MBI. 2-Inclusion criterion was that educational stakeholder (teacher, administrator, supervisor and academician) was chosen as the sample of the study. 3-The study data included statistics that could be transformed in an effect size. 4-The study was available in Turkish or English. 5-The sample was from Turkey only. Final database consisted of 75 studies, which yielded 75 effect sizes for each dimension of burnout inventory, total 225 effect sizes representing 19.827 participants, 9.017 females and 10.810 males. All the studies measured burnout using the MBI-ES or MBI-HSS (Maslach et al. 1996). Four methods were used to test publication bias: Funnel plot, Orwin" fail-safe N, Egger Regression Intercept test and Duval and Tweedie"s trim and fill method. These four statistics were chosen because they are widely used and easily understood. The mean effect size EE dimension is 0,08 (p<.05) with a 95 % confidence interval of .030 to .139 favouring female ones. For DP dimension it is -0,05 (p<.05) with a 95 % confidence interval of -.095 to -.015 favouring male ones and for PA dimension it is 0,001 (p<.05) with a 95 % confidence interval of - .043 to .045 favouring female ones. The significant d value indicates that the obtained effect is distinguishable from zero, with females scoring somewhat higher on emotional exhaustion and reduced personal achievement compared to males. However, for depersonalization, males" mean scores are higher than female ones. The last three columns in Table 2 provide evidence that potential moderators are present in these studies. The Q statistics are all significant (QEE=207,604, p<.05; QDP=114,487, p<.05; QPA=136,784, p<.05) and I2 statistics indicate 64,35 %, 35,36 % and 45,90 % respectively, suggesting the presence of moderators, along with the potential effects of other artifacts. The results of this meta-analysis may motivate policy makers to consider how educational policies may improve the well-being of both sexes. Since burnout can cause a deterioration or depletion of emotional and cognitive resources over time, it may be important for a socio-political involvement as a part of a broader effort to combat burnout as well.