Background rnAs one of the vital events of interest death must always be certified by a person authorized byrnlaw to issue a document, stating the causes of death. In order to ensure the universal applicationrnof this principle World Health Organization has recommended an international form of medicalrncause of death certification that is adopted by most countries of the world. A good quality deathrncertificate should fulfill the criteria put forth including a correct documentation of causes ofrndeath, with the correct sequence and time intervals. In Ethiopia, however, the internationallyrnrecommended death certificate guidelines have not been utilized until now. rnObjective rnThe purpose of the study is to assess quality of facility based medical certification of cause ofrndeath in the Department of Pediatrics and Child Health at Tikur Anbessa Specialized Hospital.rnMethodsrnA hospital based cross-sectional study was conducted with retrospective data collection byrnreviewing medical death certificates of deceased patients who were admitted at the Departmentrnof Pediatrics and Child health before their death. A structured pretested questionnaire withrndetails of the deceased, WHO formatted death certificate and Rapid Assessment Tool were used.rnMedical death certificates were reviewed by the primary investigator. Errors were identified byrnthe rapid assessment tool and classified as Major and Minor errors. Disparities in the magnitudernof errors as compared to age, sex, duration of stay, time of death; qualification of certifier andrnplace of death were assessed. rnResultrnFrom the 423 reviewed medical death certificates that were reviewed, 100% of them had errors.rnThe commonest major error was a missing time interval between diseases. The commonestrnminor error was use of abbreviations. Duration of stay of less than 24 hours and certification by arnsecond year resident were associated with a higher chance of error. Death in the PediatricrnIntensive Care Unit and Hemato-oncology wards was associated with a lesser incidence ofrnerrors. rnConclusionrnAs observed by the percentage of medical death certification errors, this study confirms thatrnthere is a high magnitude of error among all of the issued medical death certificates. There is arncollectively similar quantity of poor quality certificates regardless of the attributes of therndeceased, as well as determinant factors of certifiers. It also demonstrates some of the factorsrnassociated with higher or lower proportions of errors. In addition, it has also picked thernmagnitude of ill-defined Underlying Cause of Death. These are all crucial inputs into hospitalrnand national data.