Student motivation is one proposed contributing factor to disparate rates of high school achievement. Guided by self-determination theory (SDT; Ryan & Deci, 2000a), the purpose of the present study is to investigate whether a process model of the relationship between motivation and achievement varied by gender and racial subsamples of students within one large high school (see Figure 1). First, I examined the reliability and structural validity of the measures within the model. Second, I examined the extent to which the climate of one large urban high school was related to individual perceptions of autonomy, competence, and relatedness across gender and racial/ethnic subgroups. Then, I used structural equation modeling to test a process model in which autonomous motivation mediated the relationship between perceived climate or needs-satisfaction and achievement (with and without accounting for prior year achievement). The sample consisted of 863 students from an urban high school in the Western United States. The reliability and structural validity for the Basic Psychological Needs Scale (BNS) were below acceptable ranges, whereas the reliability and structural validity for the Academic Motivation Scale (AMS) were within the acceptable ranges. The relationships among needs-satisfaction, autonomous motivation, and achievement were not fully consistent with SDT; that is, competence and relatedness were significantly related to achievement (r = .20, r = .19, respectively, p < .01) whereas autonomy was not (r = .09, p > .01). European American and Asian American students showed significantly and meaningfully higher mean-level GPA than African American and Latino students, whereas differences among groups in needs-satisfaction and autonomous motivation were not both meaningful and significant. The process model without prior achievement explained 1% of the variance in achievement and fit indices approached acceptable ranges, whereas the model with prior year achievement showed unacceptable fit indices. This study highlights the importance of valid and reliable measures of psychological constructs before examining their relationships to achievement. Additionally, the value of studying hypothesized models across gender and racial subsamples is demonstrated. Lastly, this study illuminates the importance of continually testing theories, even those that are relatively well established.