Despite the explosive growth in the number of people receiving disability benefits in South Africa, very little is known about the labour supply effects of the disability grant (DG). This study investigates the impact of disability grant receipt on labour force participation. Consideration is given to potential bias that may arise from unobserved confounding factors. The study utilises data drawn from the 2007 wave of the General Household Survey (GHS) and implements a three-step methodology in a comparative perspective. Firstly, a standard probit regression of labour force participation is applied, followed by an instrumental variable regression to correct for possible endogeneity of DG take up. Finally, the sensitivity and robustness of the results is checked by implementing a variety of propensity score matching techniques. The results overall suggest that the DG receipt has substantial work disincentive effects, but the magnitude of the effect differs between parametric and non parametric estimators.