OBJECTIVE To describe the clinical profile and outcomes of severely malnourished cases admitted at Zewditu Memorial hospital, Ethiopia. METHODS A retrospective descriptive data analysis of severely malnourished cases admitted to Zewditu Memorial Hospital from April 2005 to September 2008 was made. A total of 164 cases were enrolled and analyzed for various socio-demographic factors, comorbidities and outcomes. Both bivariate and multivariate models were performed to determine the outcome of the management by explanatory variables. Pearson's chi-square test of independence was used to test the existence of significant association of risk factors with the outcome. A p-value of less than 0.05 denoted significance in differences. RESULTS The predominant age group suffered from marasmus was the infants (75.4%) while kwashiorkor was prevalent during the second and third year and the difference noted was statistically significant. The mean age for marasmus, kwash and marasmickwash incidence was 16.9, 25.9 and 27.3 months respectively. The proportion of underweight was higher after the age of 60 months. Death occurred in 21.3% of the cases suggesting that mortality rate was higher than the acceptable range (21.3% vs. < 20.0%). Presence of diarrhoea (AOR=3.5, 95%CI=1.2 to 10.2), ocdema (AOR=0.2, 97%CI=0.1 to 0.9), stunting (AOR=3.3, 97%CI=1.2 to 8.2) and short mean duration of hospital stay (AOR=4.4 95%CI=2.0 to 10.1) were predictors of death outcome. CONCLUSIONS The observed case fatality rate is unacceptably high and the risk factors for death are identified. In the face of many shortcomings in the hospital setting, managing uncomplicated cases of severe acute malnutrition is not encouraging when compared with the promising results of community based therapeutic care. We recommend the staffs to be trained and retained.