While stocking floodplain depressions or beels with fingerlings is a common form of fisheries management in Bangladesh, bio-economic guidance for improving the outcome of stocking strategies is sparse. The Community-Based Fisheries Management (CBFM) Project, funded by the Ford Foundation and the UK Government’s Department for International Development (DfID) promoted stocking practices in beels throughout the country as a means to improve fisher livelihoods. This paper describes an empirical bio-economic model developed using data generated under the CBFM project. The model offers guidance on selecting stocking densities depending upon the available size (length) of fingerlings tomaximize profit and return on investment while minimizing risk. Because large fingerlings are relatively inexpensive and have lower rates of natural mortality, the model predicts that it is more profitable to stock large fingerlings at low densities than small fingerlings at high densities. These general recommendations were found to be largely insensitive to the market price for harvested fish. To minimize credit burden and financial risk, minimum stocking densities should be selected according to the length of fish available that maximize profit. Because of its empirical nature, the model recommendations may not be applicable beyond the project sites. Furthermore, it is recommended that attempts be made to field test the model predictions before widespread adoption or promotion.