Contemporary research on achievement motivation in physical activity settings is based largely on analyses of individual's achievement goals. Most achievement goal researchers use a performance-mastery (ego-task) goal dichotomous model to assess individual differences in goal orientation. In recent years, the dichotomous model has been developed to the more sophisticated trichotomous model (mastery goals, performance-approach goals, and performance-avoidance goals) and finally to the most recent 2 x 2 model (mastery-approach goals, mastery-avoidance goals, performance-approach goals, and performance-avoidance goals). The structure of achievement goals, however, has been a subject of controversy by many achievement goal researchers (Duda, 2001; Elliot & McGregor, 2001; Roberts, 2001). Additionally, few studies to date have examined and compared the factorial validity and internal consistency of the dichotomous, trichotomous, and 2 x 2 models simultaneously. The purpose of this study was to examine and compare the factorial validities and reliabilities of these models using a sample of undergraduates in physical activity settings. A total of 246 undergraduates (141 male, 105 female) served as participants. They completed three questionnaires: (a) Duda and Nicholls's (1992) 13-item task and ego in sport questionnaire; (b) Guan, McBride, and Xiang's (2005) 18-item trichotomous achievement goal questionnaire; and (c) Guan, Xiang, McBride, and Bruene's (2006) 12-item 2 x 2 achievement goal questionnaire. Cronbach alpha and Confirmatory Factor Analysis (CFA) were performed to examine the internal consistency and factorial validity of scores produced by dichotomous, trichotomous, and 2 x 2 models, respectively. Although Cronbach's alpha coefficients indicated acceptable internal consistency for the dichotomous model (mastery goals = .87, performance goals = .87), trichotomous model (mastery goals = .85, performance-approach goals = .82, and performance-avoidance goals = .77), and 2 x 2 model (mastery-approach goals = .77, mastery-avoidance goals = .79, performance-approach goals = .78, and performance-avoidance goals = .70), CFA analyses revealed that only the 2 x 2 model (GFI = .93, CFI = .93, TLI = .91, and RMSEA = .08) represented an adequate fit to the data. Both the trichotomous (GFI = .86, CFI = .87, TLI = .85, and RMSEA = .08) and dichotomous models (GFI = .88, CFI = .90, TLI = .88, and RMSEA = .09) did not provide a good fit to the data. The results of this study suggest that the 2 x 2 model provides a valid and reliable measure of achievement goals for undergraduates in physical activity settings compared to the dichotomous and trichotomous models.Keyword(s): assessment, college level issues, measurement/evaluation