This article proposes the estimation of the marginal cost of individual firms using semiparametric and nonparametric methods. These methods have a number of appealing features when applied to cost functions. The empirical analysis uses data from a unique sample of the California electricity industry for which we observe the actual marginal cost and estimate the marginal cost from these data. We compare the actual values of marginal cost with the estimates from semiparametric and nonparametric methods, as well as with the estimates obtained through conventional parametric methods. We show that the semiparametric and nonparametric methods produce marginal cost estimates that very closely approximate the actual. In contrast, the results from conventional parametric methods are significantly biased and provide invalid inference.