This study simulates the future impact of land use and land cover (LULC) change and the enforcement of policy bundles on the multi-functionality of tropical forest landscapes. It focusses on the Northern Ecuadorian Amazonas (NEA) which is characterised by high biodiversity and endemism but at the same time is threatened by deforestation and unsustainable land use practices. Scenario simulations mimic LULC change during 2016–2030 at canton-level (area: 2,146 km²) combining explorative with policy-screening components. Scenarios NATIONAL(+)POLICY, NATIONAL(-)POLICY, REGIONAL(+)POLICY, and REGIONAL(-)POLICY are contrasted by varying deforestation trends (high regional demand for agricultural products vs. national-level structural transformations) and the enforcement/ non-enforcement of policy bundles on forest protection, reforestation and restoration. An indicator-based impact assessment analysed in a spatially-explicit fashion for each scenario in 2030 (i) the evolution of agricultural and forestry-related LULC change trajectories, (ii) future hotspots of deforestation, (iii) remaining forest core zones (FCZ) as key element of habitat integrity, and (iv) the provisioning of long-term carbon sinks. Key findings reveal that only NATIONAL(+)POLICY, as the combination of enforced policy bundles and simulated national LULC change trend extrapolating the long-term reference period 1990–2016, lead to an expansion of forest areas and a 2% carbon stock gain in 2030 compared to 2016. When assuming a linear continuation of this trend, it requires 96 years to reach a similar carbon stock level compared to the reference year 1990. Farm area expansion will continue to affect protected forests in case of scenarios NATIONAL(-)POLICY and REGIONAL(-)POLICY but will be avoided in NATIONAL(+)POLICY and REGIONAL(+)POLICY. The simulated persistence of many farming areas reveals that the likelihood to reforest existing agricultural patches will be rather low. Against expectations, scenario NATIONAL(+)POLICY which assumes net zero deforestation by 2030 led to a high level of FCZ fragmentation contrasting its carbon sink potential (27,971 Mio. Mg ha−1) which performed best among all scenarios. Our study depicts a mismatch of future LULC change realities compared to the anticipated future achievements of the Ecuadorian land use policy framework. The simulated variability of forest change patterns (deforestation, reforestation) and farming trajectories (diversification, intensification, rotation) further suggests to foster more cross-sectoral policy approaches that acknowledges the hybrid role of local stakeholders who can be farmers and forest users at the same time. Our study shows that focusing on preserving carbon stocks and their long-term climate mitigation potential only while neglecting other important ecosystem functions or biodiversity can be a short-sighted policy goal if not accompanied by rural development programmes. We further conclude that scenario-based assessments should become an integral part of multi-stakeholder interactions across relevant levels of decision-making. This can support long-term sustainable land use management which is not only relevant for the case of the Northern Ecuadorian Amazons region but for many other landscapes in transition.