With the dramatically changing technology and market environments, the importance of technology alliance to develop new products and technologies by utilizing firms' external knowledge has increased. In order to provide insight on the relationship between teacher characteristics and technology alliance performances, this study conceptualized an alliance structure according to Lane and Lubatkin's 'dyadic construct,' consisting of student firms which absorb knowledge and teacher firms that transfer knowledge. Then we analyzed the relationship between teacher firms' relative characteristics and student firms' performance of the technology alliance, using the empirical data of the Korean IT firms that are listed on Korean stock market during 1999–2005. From this analysis, we find that teacher characteristics, such as technology capability, technology similarity, and capability for knowledge transfer, influence the performance of technology alliance.