In the upper catchments of Southeast Asia, land use change from forest to agricultural systems generated land degradation and conflicts between uplanders and lowlanders. More sustainable cropping systems are proposed to upper-catchment farmers. Grass fodder strip (GFS) is an effective anti-erosion practice, and it involves lower costs for farmers. However, labour and cash constraints are sometimes preventing farmers to implement it. To evaluate farms' current impact and adaptation capacities, we need a comprehensive understanding of farm and farm household characteristics that influence their activities. This paper proposes an approach that combines farm household surveys and modelling of farm erosion yield to help project planners and policy makers to identify such farmers in a data-scarce environment. We developed two farm typologies—one based on both farm and farm household characteristics and one based on their erosion yield and constraints. We calculated erosion yields on plot level by using revised universal soil loss equation method and identified their constraints. We found that a typology based on farm constraints and calculated farm erosion was a good complement to identify farmers who are generating the highest erosion yields and would be able to change their production systems. This methodology is mainly useful at the beginning of conservation projects, when very few hard data are available.