The adverse distributional effects of a flat tax are well-known enough and are also revealed in empirical research on Belgian data. However, advocates of the flat tax argue, correctly, that most of these studies do not take into account cost recovery. One of the important effects in that context is the possible increase in labour supply and the resulting increase in the taxable base and decrease in unemployment benefits. In this study we calculate the cost recovery of a Belgian flat tax based on a micro-simulation model that includes a labour supply model.We have found that there is indeed a clearly positive effect on labour supply and therefore on the taxable basis. By introducing a revenue-neutral flat tax, the willingness to work increases with approximately 43,000 full-time equivalents. However, the effect is limited because, compared to a static scenario, the cost recovery only allows the revenue-neutral flat tax to decrease from 39% to 38%. Furthermore, there is little or no impact of these employment effects on the strongly regressive nature of a flat tax.