Methods like Fourier descriptors and contour approximations have been investigated independently by a number of researchers for the recognition of characters. Each method has its own advantages and disadvantages. This paper illustrates that a combination of these two methods may reinforce the discriminating power of a system for the recognition of characters. In this investigation, preprocessing techniques such as filling, deleting and linking operations have been developed and applied. By making use of Fourier shape descriptors, the information generated from boundary line encodings and the inside boundary information of each input character, a set of features has been extracted for analysis. Decision is achieved by a classification scheme based on discriminant analysis and nearest neighbour classification techniques. A high recognition performance is confirmed by computer simulation and implementation of this proposed method. Experimental results obtained from 8658 samples of several data bases are briefly described. Recognition rates in the range of 99% have been obtained for hand-printed characters.