Despite several anomalies documented in the literature, the CAPM is still widely used in practice to estimate discount rates for the valuation of firms and projects. In this paper, we conduct an empirical exercise to address the errors-in-variables (EIV) problem in the estimation of the industry-level cost of equity capital. We compare the results obtained using naive OLS against those from an instrumental variables approach developed by Dagenais and Dagenais (1997). Although the presence of EIV is statistically not detectable in some cases, we argue that the differences in the output from the two estimation techniques can lead to considerable gaps when it comes to use the results in any discounted cash-flow analysis.