The purpose of the present study was to evaluate the development andrnvalidation of an Entrance Examination which has been employed for selectingrncandidate teachers for Arbaminch Teacher Training Institute.rnThe subjects of the study were 583 candidate teachers who werernselected from six zones of SNNPS in 1998/99 academic year.rnItems of the sub-tests were analyzed for 583 subjects. Reliabilityrncoefficients were calculated for the three sUb-tests by using split-half andrnKuder-Richardson techniques. A correlational and regressional analysis werernused for validating the three sub-tests, ESLCE GPA and High School averagernscore. Besides, stepwise multiple regression analysis was utilized to identifyrnvariables that explain significant amount of variance in first semester GPA.rnThe findings of the study revealed that the sub-tests English languagernand mathematics have acceptable reliability. Furthermore, the study indicatedrnthat some items from each test could be revised.rnBased on the findings of the present study it was concluded that thernEnglish language entrance examination, high school average score, andrnESLCE GPA appeared to be better predictors of first semester GPA of therncandidate teachers at the institute than mathematics and teaching aptitudernentrance examination.rnOn the basis of the findings obtained in the study, it was recommendedrnthat since Mathematics and teaching aptitude test did not contribute significantlyrnto the criterion variable, care should be made when constructing the items sornthat the reliability and predictive effectiveness of the entrance examinationrncould be improved.rnFurthermore, Utmost care needs to be taken concerning the use of thernteaching aptitude test as a selection instrument in the future. If there is thernneed to check for general teaching aptitude in candidates, expert advice shouldrnbe secured in the making of such tests