The literature on the union wage gap in South Africa is extensive, spanning a range of datasets and methodologies. There is however little consensus on the appropriate method to correct for the endogeneity of union membership or the size of the union wage gap. Furthermore, there are very few studies on the bargaining council wage premium in South Africa due to lack of data on coverage of employees under these agreements. Our study, using 2005 Labour Force Survey data, firstly reconsiders the union wage gap controlling for both firm-level and job characteristics. When correcting for endogeniety of union status through a two-stage selection model and including firm size, type of employment, and non-wage benefits, we find a much lower union wage premium for African workers in the formal sector than premia reported in some previous studies. Secondly, our study estimates bargaining council wage premia for the private and public sectors. We find that extension procedures are present in both private and public bargaining council systems, but that unions negotiate for additional gains for their members at the plant-level. Furthermore, there is some evidence that unions negotiate for awards for their members in the private sector, irrespective of bargaining council coverage.