This study examines the Capital Asset Pricing Model of Sharpe (1964) Lintner (1965) and Black (1972) as the benchmark model in the asset pricing theory. The empirical findings indicate that the Sharpe-Lintner-Black CAPM inadequately, particularly the explains Pakistan’s equity market economically and statistically significant role of market risk for the determination of expected returns. Instead of identifying more risk factors, a detailed analysis of a single risk factor is undertaken. We have concentrated on two main extensions of the standard CAPM model. First, the standard model is extended by taking higher moments into account. Second, the risk factors are allowed to vary over time in the autoregressive process. The result of unconditional non-linear generalisation of the standard model reveals that in the higher-moment CAPM model the investors are rewarded for co-skewness risk. However, the test provides marginal support for rewards of the co-kurtosis risk. Finally, the empirical usefulness of conditional higher moments in explaining the cross-section of asset return is investigated. The results indicate that the conditional co-skewness is an important determinant of asset pricing, and the asset pricing relationship varies through time. The conditional covariance and the conditional co-kurtosis explain the asset price relationship in a limited way. It is concluded that Kraus and Litzenberger (1976) attempts to develop a modified form of the SharpeLintner-Black CAPM and is more successful with KSE data.