Although a growing theoretical literature points to credit constraints as an important source of inefficiently high child labor, little work has been done to assess its empirical relevance. Using panel data from Tanzania, the authors find that households respond to transitory income shocks by increasing child labor, but that the extent to which child labor is used as a buffer is lower when households have access to credit. These findings contribute to the empirical literature on the permanent income hypothesis by showing that credit-constrained households actively use child labor to smooth their income. Moreover, they highlight a potentially important determinant of child labor and, as a result, a mechanism that can be used to tackle it.