It is well accepted that biological innovations, particularly varietal improvements, have greatly contributed to agricultural yield and output growth in the past. At the same time, public funding for breeding programmes such as at the Agricultural Research Council in South Africa has dwindled. In an effort to confirm the importance of continued funding of varietal improvement programmes, this paper estimates the benefits from wheat varietal innovations and attributes them to the different institutional sources (public, private and others) that have contributed to varietal changes in South Africa. The empirical analyses used data on market shares of wheat varieties planted by farmers and annual quantities of wheat produced across different wheat-production areas in South Africa (summer dryland, dryland winter, and irrigation). A vintage regression model was estimated to calculate the proportional yield gain from wheat varietal improvements. The results indicated that the rate of gain in yield as a result of releases of new wheat varieties (variety research) was 0.8 per cent per year (equivalent to 19.84 kg/ha/year) for dryland summer varieties, and 0.5 per cent for both irrigation (equivalent to 32.20 kg/ha/year) and dryland winter varieties (equivalent to 16.65 kg/ha/year). The attribution of benefits among different institutional sources confirms that not accounting for attribution of benefits by source and time period results is overestimation of benefits to any specific research programme. Attribution of benefits by institutional source showed that Sensako dominated, while the share of the ARC-SGI substantially declined, after deregulation of the wheat sub-sector. The results highlight the impact of the decline in public funding for wheat variety improvement research after deregulation and provide a strong argument for continued public funding for variety improvement in South Africa.