Zimbabwe's CAMPFIRE program was an influential early effort to implement community-based wildlife management on a national scale. From its inception in 1989 through 2000, international donors provided substantial support, and researchers studied outcomes closely. In 2000, local CAMPFIRE projects began to feel the impact of two powerful external shocks: the end of international funding, and the beginning of Zimbabwe's severe political and economic crisis. These shocks created natural experiments in which hypotheses regarding variables affecting project outcomes can be tested. This article reports the results of case studies conducted in 2004 and 2006 of two previously highly regarded local projects. We found, in line with expectations, that quality of governance and community benefits in both sites have declined sharply. We also found, contrary to expectations, that project revenues and conservation benefits have been largely sustained. We discuss the implications of these results for theory and practice in community-based wildlife management.