Most studies on the economies of scale and scope in the banking industry assume no x-inefficiencies. That is, banks are assumed to be always on their efficient frontier, which can in empirical studies confound scale and scope efficiencies with x-efficiencies. The current paper employs a stochastic frontier cost function incorporating technical and allocative inefficiencies, as well as a system of share equations, to estimate scale and scope efficiencies. Using data from Taiwan's banking industry, evidence is found that both scale and scope economies exist, and that the assumption of no x-inefficiencies results in underestimating such economies.