Coaches have great influence on their teams, and their leadership styles and behaviors have a great effect on the performance of their athletes. The aim of this study was to investigate the relationship between coaches’ leadership styles and group cohesion in selected student teams in Urmia city. The sample of the study consisted of 75 high school students who completed two questionnaires of this study: leadership scale for sport (LSS Chelladuria & Saleh, 1980) and group environment questionnaire (GEQ, Carron, Widmeyer & Brawley, 1985). LSS contains 40 items that measure five dimensions of leadership styles (training and instruction, democratic behavior, autocratic behavior, social support and positive feedback. The GEQ with 18 items assessed two dimensions of task cohesion and social cohesion. Cronbach’s alpha coefficient was utilized to examine the internal reliability of LSS (r= 0.82) and GEQ (r=0.81). Data were analyzed with one-sample Kolmogorov-Smirnov, Pearson correlation coefficient, ANOVA with repeated measures and Bonfferoni post hoc test at P<0.05. The results showed that coaches used training and instruction leadership style more and democratic style less than other styles. There was a positive significant relationship between social support (r=0.49; r=0.37), training and instruction (r=0.56, r=0.38) and positive feedback (r=0.34; r=0.33) with task and social cohesion and a negative significant relationship between autocratic style (r= -0.17, r= -0.18) with task and social cohesion. The results of this study and previous studies showed that coaches’ leadership styles and behaviors had a great effect on team cohesion and coaches play an important role in developing team cohesion.