Previous research finds that some environmental benefits stemming from the Conservation Reserve Program (CRP) are offset by slippage: farmers simply plant more acreage to substitute for land that was idled. Our analysis shows that previous slippage estimates likely stem from spurious correlation. Most land retired under CRP is of lower-than-average quality. Due to the marginal economic viability of these lands they also are more likely to move both into and out of agricultural production. CRP enrollments therefore will be spatially correlated with non-cropland to cropland conversions even if no slippage is present. Using time-series rather than cross-sectional variation in CRP enrollments, we obtain new slippage estimates that control for land heterogeneity using fixed and random effects. Contrary to previous findings, we find little or no slippage in the form of new plantings of commodity crops. Moderate CRP-induced plantings take the form of new hay plantings that arise mostly from converted pastureland, but these conversions create little in the way of unintended environmental damages. Total commodity production is reduced by less than the proportion of acres idled because land retired is of lower-than-average quality and because it sometimes stood fallow or in hay before it was enrolled in CRP. Aside from its policy implications, our study demonstrates the crucial importance of accounting for spatial heterogeneity in empirical research.