Vulnerability is the probability that a household will fall into poverty in the future and a household is vulnerable if it is expected to be poor in the near future. Vulnerability is an ex ante (forward- looking) concept where poverty is an ex post concept. Bangladesh is very vulnerable to disasters because of its geography, population density and extreme poverty. For more than a decade now, national poverty assessments have been used on a routine basis to inform policy discussions on poverty alleviation in developing economy. Vulnerability measure can also help a person to take risk mitigation strategies. For the analysis we used the data of the Household Income and Expenditure Survey 2000 and 2005’s of BBS (Bangladesh Bureau of Statistics). By simple random sampling, we collected 500 households from rural households. In this study, we have used Vulnerability to Expected to Poverty, Vulnerability through Bayesian concept and Vulnerability through Bootstrap method. All these methods give different results and natural conjugate method shows efficient estimator compare to Feasible Generalized Least Square, non-informative prior and bootstrap methods. The most of the vulnerable obtained in vulnerability using non-informative prior and bootstrap methods. In the year 2005, estimated vulnerability rate is higher than the year 2000. The result of this study indicates that agriculture headed and female headed households are more poor and vulnerable. We also find that educated and aged headed households are less poor and vulnerable.