This study examines the bachelor's degree attainment among students who first matriculate in four-year institutions but who then transfer to other four-year institutions. It constitutes a contribution to the research that has primarily studied transferring students who start at two-year colleges. This study uses Heckman's two-step model to correct the sample selection bias. The major finding of this study is that transfer students, especially those with broken enrollment, have a much lower probability of bachelor's degree attainment than students with unbroken enrollment. Possible policies to improve the efficiency of four-to-four transfer are also discussed. (Contains 8 tables and 1 footnote.)