We report the results of a randomized field experiment designed to stimulate savings by migrant workers and remittance recipients. We offered U.S.-based Salvadoran migrants the opportunity to open savings accounts with a partner bank in El Salvador, and tracked savings using internal data of the partner bank alongside detailed surveys of migrants and their primary remittance recipients. Migrants were randomly allocated to being offered accounts that varied in the migrant's ability to be an official account “owner”. We find positive effects of savings when migrants are offered accounts that allow them joint ownership with remittance recipients. We also find evidence that migrants seek control over savings: those in the treatment condition that offered migrants exclusive control over bank accounts allocated more funds to accounts in their own names. While effects on savings at our partner bank were substantial, there was a similarly large increase in savings outside of our partner bank (including U.S.-based banks). We interpret this as due to the financial advice we offered as part of the treatments: migrants implemented savings strategies suggested by us but using savings facilities at other banks. We document very large treatment effects: compared to a base of roughly $800 in reported comparison group savings, offering joint or exclusive control of bank accounts leads total savings in the combined trans-national household (migrant plus remittance recipient) to increase by 96-136%.