The aim of this paper is to examine the extent to which the decline in poverty in Bangladesh and Pakistan can be explained by the manifold increases in remittances in both countries and, more importantly, the mechanism through which this worked. The methodology used is not based on sophisticated economic modelling or growth accounting and while some regression analysis is undertaken, its basic approach is to identify the key economic factors that can explain this decline. The basic conclusions are two-fold. The first that there is strong evidence that remittances do not directly flow to either the poor or the poorest households and the main mechanism through which poverty was reduced in both countries is its indirect effect through generating jobs and incomes both overall but more so at the local or district level. Second, the evidence shows that Bangladesh was able to achieve sustained high economic growth in this period as a result of better macroeconomic management which remittances, by easing the foreign exchange constraint, made possible as compared to Pakistan which despite similar increases in remittances was unable to do so. In this case, the much better export performance of Bangladesh (mainly readymade garment) as compared to Pakistan was also an important factor besides remittances.