This paper studies the role of technology in local-government tax collection capacity in the developing world. We first conduct a new census of all local governments in Ghana to document a strong association between technology use and property tax billing, collection and enforcement. We then randomize the use of a new revenue col- lection technology within one large municipal government. Revenue collectors using the new technology delivered 27 percent more bills and collected 103 percent more tax revenues than control collectors. Collectors using the new technology learned faster about which households in their assigned areas were willing and able to make payments. We reconcile these experimental findings in a simple Beckerian time-use model in which technology allows revenue collectors to better allocate their time towards households that are the most likely to comply with taxpaying duties. The model’s predictions are consistent with experimental evidence showing that treat- ment collectors are more likely to target households with greater liquidity, income, awareness of taxpaying duties, and satisfaction with local public goods provision.