CONTEXT: The semi-arid zone of Burkina Faso is characterized by strong climate variability and declining soil fertility associated with low biomass production. OBJECTIVE: The main objective of this study was to identify key levers to improve biomass management in semi-arid Burkina Faso for diverse farm types. METHODS: Farm diversity was captured with a statistical typology complemented by a participatory typology established with local farmers. Biomass management was described for the different farm types based on survey information obtained from 228 households across two villages. Fuzzy Cognitive Mapping (FCM) was conducted to represent biomass production strategies of each farm type. After a sensitivity analysis which revealed model robustness, scenario analysis was performed with the FCMs to explore farm type-specific options for alleviating biomass scarcity. Two contrasting scenarios were built based on observations and insights from the survey and focus group discussions with farmers and included (1) deliberate exchange of crop residue with manure, and (2) a subsidies policy allowing a reduction in prices of 30% for farm inputs coupled with increased off-farm revenue for the subsistence-oriented farms. RESULTS AND CONCLUSIONS: The statistical typology identified four farm types, distinguishing between subsistence or market-oriented farms and crop or livestock-oriented production. The participatory typology partly confirmed these four main types, even though other criteria of distinction were given by farmers. Different farm types used contrasted strategies in biomass production and management which were mainly driven by the total cultivated area and the herd size. The farm type with the largest herd and smallest cultivated land was the only one to rely on grain inflow from outside the farm to meet its household food requirement. The inflow of crop residue was also largest for this farm type. In contrast, crop residue outflow was mainly observed for the subsiste