A number of government programs, including USDA conservation programs, provide financial incentives to entice changes in behavior. An important question for these programs is whether temporary payments can lead to persistent behavioral changes. Over the past 20 years, the USDA Environmental Quality Incentives Program (EQIP) has provided more than $250 million to farmers adopting no-till crop production. In contrast to conventional tillage, which turns over the soil prior to planting, no-till can produce a number of environmental goods such as soil carbon sequestration, especially if farmers adopt no-till continuously for a long time period. This study examines whether temporary no-till payments result in persistent adoption of no-till beyond the term of conservation contracts. In the first part of our analysis, we examine field-level survey data, model no-till adoption as a second-order Markov process, and establish that in general there is considerable persistence in farmers’ tillage decisions. In the second part of our analysis, we examine a unique dataset of satellite-based estimates of field-level residue estimates in the Northern High Plains and examine changes in residue before, during, and after enrollment in EQIP. We conclude by discussing the potential implications of persistence for program outcomes as well as the challenges in identifying the mechanisms driving persistence.