In this study demand relations for primary oilseeds in South Africa is estimated and interpreted with the use of econometric models. Two different models, namely the Linear Approximate Almost Ideal Demand System (LA/AIDS) and the two-step Error Correction Model (ECM), were applied to annual oilseed data for the years 1971-2002. The F ratio test for separability failed to reject the null hypothesis of weak separability in most cases, indicating that sunflower seed, soybeans, groundnuts and cotton could be included in the same system and modeled together. The Hausman test for exogeneity was conducted and proved that the expenditure variable included in the estimated equations is indeed exogenous. The exogeneity of the expenditure variable provides assurance that the Restricted Seemingly Unrelated Regression (RSUR) method of estimation will provide efficient parameter estimates. Both the short run models are estimated in differenced form, from where the parameter estimates obtained were used to calculate compensated, uncompensated and expenditure elasticities of demand.