Retrieval of the stored information is a key issue. Especially image retrieval needsrnan emphasis , because the nature of the data is complex and difficult to retrieve .rnThere are many problems to be studied in the area of image retrieval. From these ,rnDocument Image Retrieval is one of the issues that have to be given attentionrnDocument retrieval can use either a textual-based retrieval system or animate basedrnretrieval system. Document image retrieval system can also be done in twornways: recognition-based document image retrieval or document image retrievalrnwithout explicit recognitionrnrnCurrently , little has been done on the Amharic document retrieval systems . ThernAmharic text retrieval systems which are covered by the researchers consideredrnlimited Amharic documents that are available only in hardcopy formatrnThe proposed system incorporates document images and user queries . Therndocument image is preprocessed , segmented at word le vela and the feature of eachrnword is extracted . Then the textual query is rendered to convert into an imagernquery, preprocessed , segmented and the feature is extracted . The technique usedrnfor feature extraction considers the word shape analysis . The extracted feature of thernimage query is matched with the feature of the document images , at word level usingrnEuclidean and cosine similarity measures . Finally relevant document images arernretrieved in ranked order in response to the given query.rnTo verify the validity of the approach proposed , experiment is carried out on 121rnscanned Amharic documents that are selected from printed legal documents andrnnews items.rnThe data retrieval effectiveness is measured using retrieval measures such asrnprecision , recall and F-Score .rnThe experimental results confirmed the validity of the model for retrieving relevantrndocument images from the collection of scanned document images.