Background:Indicators are developed to measure the feeding practice of a child. WHO hasrndeveloped an indicator to asses exclusive breast feeding which is obtained by using single 24rnhour recall. Single 24 hour recall only captures the current status. A single day dietary historyrnmay be misleading in determining a usual intake of an infant if there is a day to day variation inrnthe feeding pattern, and this might lead to an over estimation and misclassification.rnObjective:To assess how accurate a single 24 hour dietary recall, multiple 24 hours recall andrnrecall since birth is as compared to 7 repeated 24 hour recall in assessing exclusive breastfeedingrnamong infants less than 6 months, Butajira, Ethiopia.rnMethod:Community based cross sectional study was conducted from March to April 2016. Arntotal of 422infant mother pairsless than 6 months who resides in Butajira were selected by simplernrandom sampling method. Data on duration of EBF was collected by using a single 24 hourrnrecall, multiple 24 hour recall and recall since birth.Mcnemarâ€™s test was done to asses if therernwas a significant difference in rate of EBF. P-value less than 0.05 was considered to indicate arnstatistically significant difference in proportion. Sensitivity, specificity, positive predictive valuernand negative predictive value were computed by using 7 repeated 24 hour recall as a reference.rnResult: exclusive breast feeding rate varied across different methods.The highest prevalencern76.7% was obtained by single 24 hour recall. 7 repeated 24 hour recall and recall since birthrnresulted in EBF rate of 53.2% and 50.2% respectively. Single 24 hour recall overestimates EBFrnprevalence by 23.54%;by increasing the number of recall we can decrease the degree ofrnoverestimation significantly. Taking seven repeated 24 hour recall as a gold standard, single 24rnhour recall was observed to have the lowest specificity of 49.7% and positive predictive value ofrn69.3%. An increase in specificity as high as 94.8% was observed by increasing the number ofrnobservation days. Recall since birth was found to have a high specificity 93.8% and positivernpredictive value 94.2.rnConclusion:Single 24 hour recall overestimated EBF prevalence and had the lowest specificity.rnBy increasing observation days we can improve accuracy of estimates of EBF. Recall since birthrnpresented estimates of EBF that is close to reality. The use of recall since birth could be arnfeasible alternative to assess EBF practice.