In this paper we shall be analyzing gradual poverty reducing reforms for monetary transfers in Belarus. Some evidence is offered in support of the claim that the effectiveness of the existing 'categorical' system of transfers might be improved by redirecting resources to specific groups of households. Using a decomposition of households into socio-economic groups, feasible marginal reforms are suggested, and their robustness is checked by both experimenting with different poverty lines and introducing a simplified form of labour supply responsiveness. Based on these findings, a microsimulation of a hypothetical transfer change among different groups of households is run. According to this microsimulation, all aggregate poverty indicators could be reduced.