Globally, there is an emerging interest in scaling the adoption of land restoration interventions as a sustainable approach to improving food security. Past attempts to scale restoration interventions have tended to promote a few generic options that are often poorly tailored to their context. This has resulted in restoration interventions that are not locally adapted, acceptable or cost-effective. I explored the potential for utilising local knowledge to better inform land restoration options across scale. Local knowledge was elicited through a systematic knowledge-based systems approach involving smallholder farmers (n=482) using ‘paired-landscape experimental design’ for degrading and recovering systems in both Rwanda and Ethiopia. My findings demonstrate that in landscapes affected by degradation, livelihood systems operate across broad landscape scales beyond the farm boundary and have wide spatial livelihood system boundaries and as a result the thesis was able to capture locally informed scaling dimensions. This included identifying use of indicators, such as Dichogaster itoliensis (an earthworm new to science) whose burrowing behaviour is a local indicator of soil quality. Local knowledge can also be used to inform understanding about ecosystem service scaling processes (both spatial and temporal) but also for identifying critical knowledge gaps in farmer understanding about degradation processes. This local knowledge informs how farmers adapt and modify their land restoration interventions to better suit their needs and context; hence the acquisition and analysis of local knowledge provides an effective mechanism to track iterative development of adaptation measures and to evaluate both positive and negative consequences resulting from these actions. These findings support the ‘options by context’ approach to ‘research in development’ for adapting restoration technologies (with a focus here on agroforestry systems) to better suit the needs of smallholder farmers trying to recover their soils and points towards the need for further integration of local knowledge in the development of restoration activity.