Targeting the ultra poor (TUP) is an integrated, grant-based, anti-poverty intervention in rural Bangladesh. It combines productive asset grants (mainly livestock), confidence building, enterprise training, healthcare, cash and social security support for protecting the ultra poorâs consumption, promoting their wealth accumulation and preventing the negative effects of illness. This thesis examines TUPâs effects on health (i.e., physical, behavioural, psychological and promotional) and well-being (i.e., food consumption) of ultra poor people. It analyses the impact dynamics (i.e., short-, medium- and long-terms) of health and well-being, food consumption vulnerability to shocks and TUPâs differential effects on well-being. The thesis examines these issues constructing a more reliable matched panel from the BRACâs original balanced panel dataset, which is quasi-experimental in nature. The estimates using conditional difference-in-difference (DID) approach with household fixed effects indicate that TUP has had beneficial effects on health and well-being including reduced illness, healthcare sought from modern practitioners, self-reported health improvements, clean water, safer sanitation and improvements in overall wellbeing. However, these health outcomes are not sustained in the long-term (even decay) because of the program design and the lack of household preferences to maintain and invest in health outcomes (e.g. sanitation). In contrast, findings confirm the long-term beneficial effects on well-being. It also suggests the use of BRACâs original sample may lead to a downward bias in TUPâs impact assessment on health and well-being as it does not adequately control for the differences in initial conditions among the treatment and control households. Analysis reveals that both health and non-health shocks are prevalent to ultra poor households and are associated with their food consumption vulnerability. However, TUP smooths food consumption through direct and indirect mitigating effects. The results also suggest that being a TUP household improves access to NGOs for shock-induced borrowing. However, food consumption insurance role does come at the cost of production efficiency. Further empirical investigation using the quantile treatment effects (QTE) and the conditional quantile difference-in-difference (QDID) approaches point to a minor differential effect on well-being in the medium- and long-terms. Finally, the QDID approach shows that though TUP works best for the poorest of the ultra poor, it impacts well-being positively from the lower to the upper tail of the food consumption distribution. Therefore, estimation methods used to the matched panel confirm a positive and lasting impact of TUP on food consumption, which supports the extreme poverty-alleviating effects of TUP in rural Bangladesh.