Comprehensive school reform (CSR), a federally funded program, is designed to raise student achievement via the implementation of whole school reforms. Unfortunately, although close to 400 CSR models have been adopted by schools nationally, there is relatively limited empirical evidence from rigorous research studies regarding the effectiveness of CSR in general or of specific models in particular. Of additional importance is the extent to which CSR models can demonstrate positive effects over time. The purpose of this study is to evaluate the effects on student achievement and attendance, in an urban school district in Kentucky, of the Year 3 implementation of the Different Ways of Knowing (DWoK) for the Middle Grades model. Results indicated that achievement results on both the Comprehensive Test of Basic Skills and Kentucky Core Content Test were significantly superior for the DWoK schools relative to matched control schools; no positive impacts on student attendance occurred for Year 3. The implications of the study lies in revealing that learner-centered CSR models that integrate the arts with the core curriculum can produce observable achievement gains in both normed- and criterion-referenced tests. (Contains 1 footnote and 6 tables.)