The Forest Department of Bangladesh initiated agroforestry program for the poor people in order to alleviate poverty. This study explored to what extent the agroforestry program had reduced the poverty and which constraints might be responsible for poverty alleviation among the participants. The research made use of Head Count Index (HCI) to determine to what extent poverty was decreased. To understand the main constraints of agroforestry program, this study focused on macro-meso-micro level analysis technique. Data were collected through semi- structured questionnaires for face-to-face interviews within the study area i.e., Ghatail upazila (sub-district) of Tangail District, Bangladesh. Total fifty agroforestry participants were randomly selected form the study for this research study. The results showed that this program had alleviated poverty at a significant level (64% of the respondents) which improved their livelihood situation considerably. The income of the settlers in a similar participatory forestry program varied due to their socio-economic factors such as months of food sufficiency, family size, distance to market and institutional loan. These were the main reasons for the income differences among the settlers. As regards the constraints as perceived by the three level analysis frames, bureaucracy, monopoly of the market system, poor road infrastructure and the lack of loan facilities were considered to be the main problems to reduce poverty with this program. Considering the overall results of the study it can be concluded that the Participatory Agroforestry Program (PAP) was quite successful for increasing income as well as alleviating poverty. If we can overcome these constraints then there are better possibilities of agroforestry to contribute better to alleviate poverty situation of this study area as well as Bangladesh.