The general goal of this paper is to analyze poverty in household per capita consumption as a monetary measurement and based in the data from Albania trying to identify probable determinants that influence in falling in a trap of poverty. Current literature suggests several ways of modeling the determinants of poverty. Usually the regression analysis is used to check in the same time the influence of the different factors. In this paper, binary logistic regression was estimated with economic status (poor-non poor) as dependent variable and a set of characteristics of individual and household as independents variables. The logistic model used shows that probability of being poor is found to be influenced mainly by education and status of employment of household head, the household composition and geographic divisions.