Addressing the complex links between the economic and ecological sphere, my dissertation seeks to shed a light on the socio-economic drivers of land-use change in developing countries. Taking Indonesia as an example, this thesis will specifically focus on the determinants and impacts of land-use change that are relevant for income growth of farm households in developing economies. The first analysis uses a meta-analysis to explore the underlying drivers of land-use change at the farm household level across the tropics. In reviewing 91 recent empirical and theoretical studies, this chapter identifies the key determinants of households’ decisions on land-use change within the current literature. The literature on micro-level drivers on land-use change points towards micro-level economic growth (e.g. in income and capital endowments) being a strong catalyst of human-induced land-use change. Further, the review suggests that there is substantial heterogeneity among farm households regarding these endowments, which is also significantly associated with households’ land-use changes. The second paper explores how cocoa cultivation contributes to poverty reduction and whether income gains from cash crops are more volatile. The results show that cocoa cultivation is associated with strong and sustainable poverty reduction. Yet, yield gaps among cocoa smallholders remain large and are increasingly heterogeneous. Productivity heterogeneity can be traced back to farm management practices. Linking these findings to poverty transitions, results show that better management practices facilitate the transition out of poverty and shield against income losses. Having gained insights into households’ potential of cash crop cultivation, the third paper presents a dynamic ecological-economic model of land-use change. The integrated model explores the potential of landscapes with different land-use patterns to balance ecological and socio-economic goals. Within the economic submodel, smallholders’ land use and management decisions are based on a profit maximization assumption bounded by the available wealth of that household. Households’ land decisions are directly linked to the ecological submodel, which includes a simple account of carbon sequestration in aboveground and belowground vegetation. First simulations show that the relationship between carbon accumulation/storage and economic benefit might not be completely straightforward. The fourth paper complements the previous papers’ focus on the micro-level determinants of land-use change by concentrating on the broader scale effects, particularly on the trade-offs between economic gains and the loss of ecosystem functions achieved through the agricultural specialization within transformed landscapes. The analysis takes Jambi province, Indonesia, a current hotspot of rubber and oil palm monoculture, as a case study to illustrate these issues. It empirically shows that the level of specialization differs across scales, with higher specialization occurring at the household and village levels. Findings further suggest that there are gains from specialization at the farm level but that this specialization does not necessarily lead to a consolidation of smallholder farms to ever-larger units. This result can be set in the context of a conciliating landscape design within multi-functional landscapes, where land use patches of highly specialized smallholders are intermingled with areas characterized with high levels of ecosystem services.