This chapter opens by introducing key issues in stochastic frontier analysis, including the wrong skewness problem and the difficulty of interpreting estimated inefficiency as evidence of organizational inefficiency. Based on these considerations, the chapter then clarifies the purpose of the book and its basic theoretical and econometric approach. It presents two theoretical frameworks for analyzing technical inefficiency, a vintage model and a model that assumes the presence of important unobserved inputs. The chapter also explains how the implications of these models are connected to observable relationships among residual moments across many industries and outlines the estimation strategy used throughout the book.