When 70 percent of the world's poor are rural with agriculture as their main source of income, the intersection between land rights and land use becomes increasingly important to global poverty reduction. I explore the ways by which dimensions of land rights shape and distort land use away from its optimal allocation. Using a wide range of data, including satellite imagery, censuses, and household surveys, I present empirical evidence that demonstrates the effect of ill-defined and limited property rights on land use in poor, rural communities in two diverse countries. Well-defined private property rights over land should incentivize efficient transactions between owners, efficient levels of investment as well as optimal allocation across different uses. Thus a strengthening of property rights would result in a land allocation closer to the counterfactual private property outcome due to a reallocation of land across uses (depending on their relative returns). In the first chapter, I examine the impacts of part of Mexico's second agrarian reform in 1993, Programa de Certificacion de Derechos Ejidales y Titulacion de Solares (Procede), which certified all land in Mexico's ejido communities. Using LANDSAT images of ejido and non-ejido land to characterize land use and suitability for different uses in Mexico over this period, we find that the average ejido does in fact alter its allocation of land across forest, agriculture and pasture in response to certification. While the average results indicate that Procede had a positive effect on forest (31 ha.), an offsetting negative effect on pasture (29 ha.), and no effect on agriculture, we explore further heterogeneity based on estimated land suitability. Using several spatial datasets of physical, climatic and economic characteristics, we estimate land suitability based on private property, non-ejido land in Mexico. The pattern suggests that strengthening property rights induced a convergence of ejido land allocation to the allocation implemented under private property. In total, the area deforested over 1990-2010 would have been approximately 14 percent higher, there would have been 40 thousand fewer hectares of cultivated land, and 715 thousand more hectares of pasture had Procede not been implemented and ejido land left uncertified. The next chapter focuses instead on Vietnam, in which the state takes advantage of incomplete property rights to directly influence land allocation decisions. Across many economic contexts, there are policies whose efficacy is undermined by endogenous responses of agents due to a misalignment of incentives. In this chapter, I show that households' production responses to a food security policy in Vietnam that restricts household land to be used for rice considerably undermines the policy's purpose. I develop a model of farmer crop choice that demonstrates how divergence of interest between the farmer and commune authority, and subsistence rice production constraints for the household generate different testable predictions for the impact of restrictions at both the household and plot levels. I test these predictions using four rounds of the household and plot panels of the Vietnam Access to Resources Household Survey (VARHS) between 2006-2012. The evidence suggests that land use restrictions are largely ineffective at increasing household rice production and lower agricultural profits. This is due to the fact that households reduce rice production on their unrestricted land while complying with restrictions. Counterfactual household rice production without any such 'slippage' on unrestricted land is 12 percent higher, and I estimate that restrictions reduces household agricultural profits by 15 percent on average. Thus, the policy appears to be unsuccessful in increasing household rice production while at the same time imposing welfare costs to the household.