This paper examines whether offering landless forest-adjacent communities options to grow appropriate food crops inside forest reserves during early stages of reforestation programmes increases incomes of low-income households and conserve forests. We consider the forest cover and household welfare impacts of a unique incentive scheme in Kenya known as the Plantation Establishment and Livelihood Improvement Scheme (PELIS). PELIS seeks to deepen community participation in forestry, and improve the livelihoods of adjacent communities. Using cross sectional data collected from 22 Community Forest Associations and 406 households, we use propensity score matching methods to evaluate the mean impact of the scheme on forest cover and household welfare. We also assess the heterogeneous impacts of the scheme on household welfare using an endogenous quantile treatment effects model. The results show that on average, PELIS has a significant and positive impact on the welfare of participating households (estimated between 15.09% and 28.14%) and on forest cover (between 5.53% and 7.94%). However, the scheme cannot be defended on equity grounds as it has inequitable distributional impacts on household welfare. The scheme raises welfare of groups other than the poorest and marginalized sections of the community. Our observations from the field blame elite capture for this outcome.