The primary purpose of this research was to investigate the relationships between Kelly's (1955) personal construct system and the degree of career development for junior high school students. Two studies were conducted via quality live method and quantitative method respectively for this purpose. In study one, the revised career choice repertory grid and the Career Development Inventory were administered to 426 subjects. The data were analyzed by canonical correlation. In study two, eight types of the participants were categorized by the degree of differentiation, integration and conflict ratio: Primitivity A, Primitivity B, preemption A, ptreemption B, Circumspection A, Circumspection B, Control A, Control B. The participants were selected from study one and then accepted interview. The transcripts of interview were analyzed by the Hycner's (1985) phenomenological method. The major findings were as follows: (a) in study one, the data indicated that the relationships between career construct system and the degree of career development turned out to be unlinear for junior high school students. The degree of integration had positive relations with the degree of career development, however, the conflict ratio had negative relations with the degree of career development, and the degree of differentiation had inconsistent relations with the degree of career development; (b) in study two, the Primitivity type held lower degree of career development and the Control type possessed higher degree of career development as predicted by the theory, while the other types were somewhat vague in between. The implications of the findings for the personal construct theory and career counseling were suggested.