The authors investigated predictive relationships among student characteristics that influence motivation for learning and achievement. Participants were students (N = 900) from all 4 grade levels in 18 rural public high schools in the southwestern United States. The authors used AMOS 4.0 (J. L. Arbuckle & W. Wothke, 1996) to test a hypothesized path model in development and validation subsamples. Variables included perceptions of classroom climate, perceived ability, perceived instrumentality of instruction, and achievement goals as predictors of engagement and effort in school. The model fit the data reasonably well, with relatively minor variations in the strength of the paths between subsamples. The authors discuss implications for research and classroom practice. (Contains 3 figures, 4 tables and 3 notes.)