The purpose of this study was to determine to what extent teacher-maderntests are effective and to identify the factors that inhibit high schoolrnteachers of Addis Ababa from analyzing and evaluating their exams a fterrnthey have administered them. To check the quality of the test, two testsrnof final exams from five sample schools were taken. This helped to checkrnthe quality of the test items from the response of the test takers byrncomputing difficulty and discrimination indices. To see the internalrnconsistency of the tests, KuderRichardson KR-20, and Spearman Brown,rns plit -h a lf methods were employed.rnMoreover, qualitative description of the test was made by observing therntest papers based on the principles of test construction . In addition, therncorrelation of average teacher-made tests of students' score and thernnational exam score of the students were computed.rnBy using qucstionnaire relevant data from teachers, and by conductingrninterview with principals about the overall condition of the schools data,rnthat can limit teachers from making posttest analysis, were secured.rnThe finding demonstrated that test qualities currently are not as suchrngood in terms of their difficulty level, discrimination power and internalrnconsistency. The qualitative data a nalyses have also indicated that most ofrnthe tests are not constructed as per the principle of test construction qrnTeachers are also found not in a position to make posttest analysis. Tornminimize the problems associated test or test items evaluation the need tornencourage teachers to en gage in such practice is highly recommended.