In this research, the authors examine whether M.B.A. students at a regional, comprehensive university who take only online courses learn as much as students taking identical courses in the traditional, face-to-face format. A simple comparison of test scores indicates that the amount of learning is similar in the two formats. However, when other factors are controlled, regression analysis shows that outcomes in the online environment are inferior to the traditional format. Given that the choice to take a course in the online environment may be related to learning, the authors use two techniques that control for sample selection bias. A two-stage least squares analysis yields findings similar to ordinary least squares estimation. However, a switching regression shows that the online learning environment is substantially less effective than the traditional learning environment. (Contains 22 endnotes.)