Educational research aims to understand how and why students change over time. With its emphasis on within-person change, latent change score models provide educational researchers with a more general and flexible framework for testing nuanced hypotheses regarding within-person change and between-person differences in within-person change. Models of change are specified and can include static (time-invariant) covariates, as well as dynamic (time-varying) covariates. Latent change score modeling further allows for the testing of more complex research questions about dynamic associations between 2 or more variables across time. We describe latent change score models for a single repeatedly measured variable and for multiple repeatedly measured variables, and we illustrate their use with longitudinal data on mathematics ability and visual motor integration. The results indicate that visual motor integration is a potential leading indicator of subsequent changes in mathematics ability.