Remittances are flows of money between immigrants and their relatives. A scarcely addressed topic related to human migration is the application of Network Flow Analysis (NFA) in the study of remittances. This study provides both a theoretical and empirical macro-level study of worldwide remittance flows. Theoretically, a remittance model used by the World Bank is probabilistically analyzed and useful probability bounds are derived. Empirically, a weighted graph model built from World Bank data delivers important insight on the remittance generative process, the highest remittance flow values and inflow–outflow remittance variations. This manuscript concludes by postulating that remittances potentially result from a multiplicative process and that techniques based on NFA provide novel ways of approaching remittance dynamics problems.