Individual household fertility has been modeled in various ways in the literature.rnIn many empirical studies of fertility, the number of children ever born is modeledrnas a function of social, economical and demographic variables such as wife'srneducation, husband's education, age atfirst marriage, child mortality ... etc. andrnthe method most commonly used is regression with OLS method of estimation.rnThere are also other studies that apply Maximum Likelihood Tobit Model. In ourrncase, i. e. for the analysis on fertility data of urban Addis Ababa both models arernadopted and may contribute to this body of research.rnThe results of the analysis evidenced that:rni) The three demographic variables - age at first marriage, desired number ofrnchildren and child mortality gave the expected results. That is, age at firstrnmarriage has a negative but insignificant effect, while child mortality and desiredrnnumber of children have marked highly significant positive effect on fertility.rnii) Virtually, consistent negative relationship was found between wife's educationrnand fertility.rniii) Wife's work status and husband education have shown no consistent andrninsignificant relationship with fertility.rnFinally, based on the findings, it is suggested that policies aimed at increasingrnwomen education, reducing child mortality and limiting desired number offamilyrnsize are useful for decreasing fertility .