Optical Character Recognition (OCR) is an area of research and development where arnsystem is made to recognize document images. Cultural considerations and enormous flood ofrndocuments motivated the development of OCR across the world. Unlike other scripts, OCRrndevelopment for Amharic characters has been started recently at SISA. Some developmentsrnhave been made in recognizing specific font styles, font sizes, and font types. But, as the fontrnstyle, size or type changes the recognition accuracy falls downrnThe purpose of this study is, therefore, to explore the possibilities of developing a versatilernOCR system that is independent of sizes of Amharic characters. To this end, differentrnpreprocessing techniques and pattern recognition techniques have been reviewed. Since thernsegmentation algorithm that was used by previous studies in the area works well, it isrnincorporated in this study with some modifications. Template matching, statistical,rnsyntactic/structural, and neural network approaches are found to be the most commonly usedrnpattern recognition techniques and the pros and cons of each technique is reviewed. To takerntheir advantage, a hybrid system of syntactic/structural and neural network approaches isrnimplemented.rnSyntactic/structural approach enables the developed OCR system to extract primitivernstructures of characters and generate a unique pattern for each character to be used by thernneural network. The neural network enables the developed OCR system to classify/recognizernthe patterns generated and it can also predict for new cases. The network takes the output ofrnthe syntactic/structural approach as an input. With this procedure, the neural network isrntrained with VG2000 Agazian font of sizes J 0 and J 2. The performance of the developedrnsystem is tested with documents written using VG2000 Agazian font of sizes 8, 12, and 14. Thernresults showed that, with minor differences, the developed OCR system classifies/recognizesrnthe test cases of different font sizes with more or less the same level of accuracy. Based on thernresults, further research areas are a/so recommended.