Objectives: Poverty in Iran is one of the interesting issues that researchers tried to find influential factors on it. However, quantitative studies about this issue, due to the lack of access to information or not using of advanced statistical methods, need to be generalized in order to specify effective factors on poverty. In this paper influential factors on poverty are determined and measured by the use of an ordinal cumulative logistic model. This model that includes the special case of binary response (poor or rich), has the advantage (relative to the available researches) that it increases states of the poverty levels to more than two categories (considering the ordinarily of the categories) and in this way, it uses more information to interpret the results. Using this model comparison of different poverty levels is also more efficient. Method: By a logical way, we have categorized the total cost variable of the urban civilians by the poverty indexes into four ordinal categories, poor, pseudo-poor, not-poor and pseudo-rich or rich. Although this categorization prevents the use of continues nature of initial cost variable, we prefer to use the created ordinal variable for two reasons. The first reason is the error due to the people’s statements and their evaluation of their total expenditure which can be reduced by the recoding the responses into being or not being in a specific category. Secondly, this categorization enables us to recognize the poverty situation of Iranian households with respect to their cost situation and calculated absolute poverty line of Iran. So, we assume that the new categorized ordinal variable which considered as an ordinal response variable takes the value 1 for poor people, 2 for pseudo-poor people, 3 for not poor people and 4 for pseudo- rich or rich people. Because of ordinal nature of the response variable, we have used the cumulative ordinal logistic model for response of poverty situation in order to find effective economical and social factors on poverty and also measuring the amount of their effects. This model can give us the ability to find the odds of being poor against the other poverty levels (pseudo-poor, non-poor and pseudo-rich or rich). By this model one can also obtain odds of being poor or Pseudo-poor against non-poor and Pseudo-rich or rich. also this model has the advantage of using the ordinal information of responses over the other simple logistic models. Findings & Results: The data extracted from the urban family's cost and income survey in statistical center of Iran, show that employment status, marriage status, education status, gender of the head of the household and also home situation and number of family members in the house are important factors which affect the probability of being poor or pseudo-poor. Results show that the chance of being poor for single household (male or female) is so less than that of households with family size of 5 or more. For households with any family size the chance of being poor for females is more than that of males. Finally, the very important result is that the poverty situation for households with high family size is critical for people who are tenants.