The basic aim of this research is to adapt the Educational Stress Scale (Sun, Dunne, Hou, & Xu, 2011) to Turkish and to examine its psychometric properties. The second purpose of this study is to compare students’ educational stress levels according to gender and grades. The research was conducted on 558 high school students in Sakarya and Agri in Turkey. Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was .81 and there was a significant result on Bartlett’s test of Sphericity χ2=3488.103 (p<.001, df=105). Factor loadings ranged from .41 to .91. Results of confirmatory factor analyses demonstrated that 16 items yielded five factors as original form and that the five-dimensional model was well fit (χ2= 123.49, df= 88 (χ2/df=1.40), RMSEA= .03, NFI= .97, CFI= .99, IFI= .99, RFI= .96, GFI= .95, SRMR= .041). Cronbach alpha internal consistency coefficient was found as .86 for whole scale, .87 for the sub-dimension of pressure from study, .93 for the sub-dimension of workload, .90 for the sub-dimension of worry about grades, .90 for the sub-dimension of self-expectation, and .91 for the sub-dimension of despondency. In the concurrent validity, significant relationship (r= .51) was found between the Educational Stress Scale and School Burnout Scale. Test-retest reliability coefficient was .67 for the whole scale. Corrected item-total correlations ranged from .40 to .57. Overall results demonstrated that the Educational Stress Scale can be deemed to be a valid and reliable instrument to be used in the field of psychology and education. Moreover, students’ educational stress levels changed statistically according to gender and grades. Results were discussed in light of the related literature.