Poverty transition through a capacity development programme called the Bangladesh Poultry Model is assessed using self-assessment dimension in a quasi experiment framework. Current poverty situation is compared with money metric measure. The programme involves longer term intervention towards building the strength of stakeholders such as government department, NGOs, village organisations and women beneficiaries. A number of key questions related to poverty transition through poultry based activities, heterogeneity in livelihood choice and its impact on household welfare, extent of poverty reduction etc. are answered for policy recommendations. Data are drawn from a survey of 400 beneficiary households in 2006; about 50% of them are survivors in the programme. Poverty profiles, transition matrices and regression analysis using asset-base framework are used to analyse data. Results are discussed along with recommendations and policy implications. Adaptation of the programme in several countries is also reviewed briefly using published information to discuss wider implications.