Since 1996, USDA working lands programs have resulted in hundreds of thousands of conservation contracts—voluntary agreements between USDA and program participants. Within each contract, the participant agrees to install or implement a set of conservation practices, and USDA agrees to provide technical and financial assistance to the participant. The use of contracts is central to the programs because of the complexity of addressing conservation goals—many contracts have multiple practices—and because of the time lags involved in installing and implementing conservation practices. A large majority of practices on these contracts are installed as planned. This study examines data on the 10 to 20 percent of practices that participants who signed contracts in the Environmental Quality Incentives Program in fiscal year 2010 did not implement as planned and dropped from their contracts. We find differences in the frequency with which specific practices are dropped, as well as in the likelihood that some types of contracts have at least one dropped practice. These differences suggest participants earn different levels of private benefits from installing or implementing conservation practices, which results in varying incentives to complete practices. Examining patterns in dropped practices reveals these private benefits, which are generally not known by program managers at the time participants sign conservation contracts. In this report, we discuss the significance of unobserved private incentives in possible policy design options—changes in ranking criteria, restrictions on contract structure, use of bundled practices at varied cost-share rates, and changes to costshare rates.