The institution of child fostering is highly prevalent in Burkina Faso despite the potential negative welfare implications for both the foster child and society. The objective of this research project is to formally model the household decision to foster a child and test several hypotheses that explain this decision. Current research is unable to explain why households would send out their children to be domestic workers. Explanations concerning labor productivity, optimal household demographic structure, or informal insurance mechanisms might resolve this puzzle. This project adds a necessary and original dimension to previous studies by using a methodology that calls for tracking households on both the sending and receiving ends of a particular fostering exchange. It will be possible to explain why those particular families exchanged a child, test the reasons the sending family fostered their child, and eliminate the measurement error bias that occurs if only one family's responses are used in the regressions. Resolving the fostering puzzle will provide information that enables development agencies to better target programs that address the underlying causes of fostering.