The empirical investment literature studying the determinants of investment relies almost exclusively on truncated samples of publicly listed firms due to the lack of data on private firms. This truncation, however, is not random because listing is a choice for many firms, whereas others cannot list due to their characteristics. The ensuing endogenous truncation entails biased estimates. We develop a methodology that corrects for the bias. The bias-corrected results lend strong support for the q-theory: the investment-cash flow sensitivity disappears and the relation between investment and q nearly quadruples. Our econometric framework is also applicable in many other economic contexts.