In accordance with the empirical regularity of time-varying betas we estimate and test for the Sharpe–Lintner CAPM by allowing for structural change(s) in betas. Empirical applications using BM- and size-sorted decile portfolios suggest the following interesting results. Firstly, there exists at least one break for all the portfolios under consideration. Secondly, the estimated break dates are quite similar for some of the portfolios, indicating the possible existence of a common break using multivariate time series. Finally, we find the CAPM can be consistent with the data in some regimes but may appear to be inconsistent with the data in some other regimes. This particularly appealing feature has been completely ruled out under the conventional single-equation framework.