A Linear Approximated Almost Ideal Demand System (LA/AIDS), estimated in first differences, was used to anticipate the demand relations for meat (beef, chicken, pork and mutton) in South Africa from 1970 - 2000. Two tests for weak separability, including an F and Likelihood ratio version, failed to reject the null hypothesis of weak separability, confirming that the four meat products are separable, and should be modelled together. According to the Hausman exogeneity test, the expenditure term in the South African meat demand model is exogenous. As a result, a Restricted Seemingly Unrelated Regression (RSUR) was used to estimate the model, whereafter the parameters were used as to calculate compensated, uncompensated and expenditure elasticities.