The purpose of this study is to look into the effect ofrneducation on migration and to identify the responses of migrantsrnwith different levels of education to the selected explanatoryrnvariables that shows costs and returns of migration. It was alsornattempted to assess the employment patterns and stability ofrnmigrants in urban areas.rnIn order to achieve the purpose of the study two phasernsampling procedure was employed to select households that were tornbe included in the study. 493 useable questionnaires wererncollected from the study town, Bahir Dar. Multiple and step wisernregression, one-way ANOVA, Chi-square and Weighted least squarern(WLS) regression were the major statistical techniques employed tornanalize the data.rnResults of the analysis show that migration rate increasesrnwith education except with migrants at the tertiary level. But therneffect is clearly seen after elementary education. Thus juniorrneducation seems to be a critical level to induce migration at thisrnlevel of the Urban hieriarchy. Results of Weighted least squarern(WLS) regression indicate that there is a marked difference betweenrnmigrants in thier responses to the factors that influencernmigration. Well educated were found more responsive to income atrnthe destination and to the unemployment period at the origin whilernilleterates were found more responsive to the period ofrnunemployment at the destination. It was also observed thatrnmigrants made a considerable shift from unemployment to state ofrnemployment and across different occupations . Further more, lessrneducated migrants were found more stable than the educated in urbanrnemployment