The sensitivity, specificity, positive and negative predictive values of a reagent strip test for the diagnosis of schistosomiasis mansoni by detecting circulating cathodic antigen (CCA) in urine were evaluated using 184 stool and urine samples collected from schoolchildren living in relatively low endemic area of schistosomiasis mansoni in Ethiopia. A combined result of stool samples processed by Kato and formol-ether concentration methods was used as gold standard. The results showed that detection of CCA in urine using reagent strip test was slightly higher than the combined results of the stool techniques (65.2 % vs 42.4 %, p > 0.05) in suggesting the prevalence of the disease. The sensitivity, specificity, positive and negative predictive values of the reagent strip test were 76.9 %, 43.4 %, 50 % and 71.9 %, respectively. The result of egg counts using Kato method suggested that detection of urine CCA could be used to indicate the intensity of infection. Nevertheless, like that of stool examination, the reagent strip test was found to be less sensitive in case of light to moderate infections. About 23.1 % of the study children who were excreting the eggs of the parasite were found negative by the reagent strip test. The relative insensitivity of a reagent strip test in low intensity of infection necessitates for the development of more sensitive assay that can truly discriminate schistosome-infected from non-infected individuals.