The Sesaot forest area has been managed by the community through the Social Forestry scheme with an agroforestry system. The objectives of the study were to identify the diversity of agroforestry patterns, analyze incomes on various agroforestry patterns, analyze the amount of carbon stock and analyze the most optimal agroforestry patterns. The method used is descriptive through a series of observations, interviews and Focus Group Discussions, to 42 respondents and measurement of carbon stocks in 18 farmer locations. Analysis of agroforestry patterns was carried out by making clusters based on plant dominance, analyzing the amount of carbon using allometric equations and analyzing the optimal value of agroforestry patterns using scoring. Based on the results of the study, it can be concluded that: there are four agroforestry patterns in the Sesaot Forest Area, namely, candlenut dominant agroforestry, mahogany dominant agroforestry, mixed agroforestry and simple agroforestry. The largest income was obtained in the mixed agroforestry pattern and the smallest was in the mahogany dominant. Meanwhile, the largest amount of carb on stock was found in the mahogany dominant and the smallest was in the simple agroforestry. The results of the scoring analysis show that mixed agroforestry practices are in the most optimal category compared to other agroforestry patterns.