This paper investigates the difficulties in measuring reservation wages, models the determinants of reservation wages, and compares reservation wages with predicted wages. Data is drawn from the Khayelitsha/Mitchell’s Plain (KMP) survey. Certain factors (e.g. education, labour market status, household income and duration of unemployment) are significant in explaining variation in reservation wages. Importantly, a person’s position in the labour market is not as a result of his/her reservation wage. Rather, reservation wages are a function of his/her labour market status: while those in wage-employment report a reservation wage based more on perceived labour market value, those in unemployment report a reservation wage influenced strongly by subsistence requirements. This study concludes that voluntary unemployment does not exist in KMP, with people in general reporting reservation wages well below what they could expect to earn.