Parental involvement in children's education is of critical importance in the U.S. educational system. Therefore, it is useful to identify effective predictors of parental involvement. The present study used multi-level analyses to examine how individual and school-level characteristics impact two forms of parental involvement (school-based and home-based parental involvement) in first grade and eighth grade. Several child/parent level characteristics significantly predicted parental involvement. Parent interaction/social capital demonstrated medium to large effects across both forms of parental involvement in both first and eighth grades. Many of the other child/parent level characteristics produced small effect sizes. Across both forms of parental involvement there were few school-level effects that were statistically significant. Those that were statistically significant were very small in magnitude. The results of the present study may serve to inform school practices and research in the field of parental involvement.