To what extent does access to banking help poor households to save and smooth consumption? To answer this fundamental question, we combine a field experiment that randomly provides access to a bank account with weekly interviews on household finances. Access to banking does not change average consumption, but it improves consumption smoothing by alleviating savings constraints. Indeed, the control’s expenditures follow income more closely than the expenditures of the treated. The latter handle variations in income by engaging in pro-cyclical saving in their account. These results provide an important new insight into the role of banking in low- and middle-income countries.