The study investigated the effects of collaborative writing (CW) on L2 development. Sixty-nine first year students in two intact classes were assigned to two groups: experimental group (n=34) and control group (n=35). In the experimental group, writing tasks were carried out in pairs; in the control group, the tasks were carried out individually. The study lasted one academic year and involved a pre- and post-test. The performances of the two groups were compared on linguistic measures of complexity, fluency and accuracy. Results showed that the language accuracy and fluency of all students significantly increased. In terms of complexity, only vocabulary sophistication improved significantly, while the lexical variation, lexical density, and syntactic complexity didn’t change significantly. In addition, the experimental group did not perform significantly better than the control group in all the aspects, suggesting CW did not have obvious advantage over traditional individual writing in promoting L2 development. The implications for carrying out CW are discussed and limitations and directions for further research are presented.