Two hypotheses are tested empirically using cross sectional data from Vaalharts: Firstly, it is hypothesized that economies of scale exist and, secondly, that optimal farm size will not be the same for any two managers; the better the manager, the larger the optimum farm size. These hypotheses are tested by means of both parametric and non-parametric methods, respectively. Managerial skill was measured explicitly. The results support both the hypotheses which were tested empirically. There also appears to be no inconsistency between the results obtained with parametric or non-parametric methods. These findings have important implications for structural adjustment of South African agriculture.