Land market incompleteness is argued to have pervasive effects in Sub-Saharan Africa, including on agricultural efficiency, equity, and structural transformation. Yet experimental evidence on land market participation is virtually non-existent. We randomly allocate subsidies for agricultural rentals in Kenya, and study who selects into land markets, what renters do differently from owners, and the resulting effects on agricultural and owner outcomes. The induced rentals increase equity - reallocating plots to farmers who own fewer plots and are younger and more market-oriented - and persist beyond the subsidy. Renters increase output and value added on the rented plot, by more than owners under an equivalent unconditional cash transfer, and they do so by increasing commercial crop cultivation and non-labor inputs, rather than labor. Although owners cultivate less land under the rental subsidy, their non-agricultural labor decreases. The results shed light on the nature and magnitude of land market frictions, and on their interactions with other missing markets.