This paper explores the regional allocation pattern of general resources (recursos ordinarios) in Peru, which are distributed to regional governments through a discretionary grant by the national government. We estimate an empirical model based on a panel of annual data between 2004 and 2010. Although national transfers are significantly biased towards regions where the national government received the lowest electoral support, the data suggest that this effect is strongest at the beginning of the administration's period in office. In the long run, however, opposition regions appear to host more volatile constituencies, which is compatible with the swing-voter hypothesis. Interestingly, regions that strongly supported the president receive the fewest benefits. Finally, the role of regional conflicts, the effect of lobbying by organized regionally based civil groups and the size of the regional constituency, among others variables, are statistically significant in the estimations.