The aim of this research is to investigate the teachers' emotional labour behaviours and to determine the reasons of the differences. In the research, mixed research methods including both quantitative and qualitative techniques were used. The population of the study was comprised of 280 teachers (266 for quantitative, 14 for qualitative dimensions) who work in Siirt city center. Because of the difficulties to reach the target population, time limitations and the cost over run, representative sampling was preferred. In order to investigate the teachers' emotional labour behaviors, "Emotional Labour Behaviors Scale" which was adapted by Kaya (2009) was administered to the teachers in the study. This scale was adapted for educational field. For the quantitative dimension of the research, 266 scales were analysed by using percentage, means, frequency and standard deviation for descriptive statistical techniques. T-test, ANOVA tests were used for the differences among the variables. As a result of the research, findings indicated that there were significant differences in the teachers' emotional labour behavior scores. The significant differences were found in the scores of the number of teachers, the working time at schools, the type of graduated faculty, and the type of the school. The questions related to the qualitative dimension of the research were asked to the teachers with different branches and gathered data were analysed using content analysis technique. According to the participant teachers opinions, the reasons for the significant differences of emotional labour behavior scores are different occupational definitions, being ready for the teaching profession, exhibiting more efforts and interest, more close relationships, and the negative effects of staying much more at school on the teachers. The participant teachers expressed their solutions as solving the teachers' time and place problems, constituting more imteractive relationships among students and colleagues and preparing effective working schedule.