The preform design in metal forging plays a key role in improving product quality, such as ensuring defect-free property and proper metal flow. In industry, preforms are generally designed by the iterative trial-and-error approach. This approach, however, leads not only to the increase of significant tool cost but also to the extended down-time of the production equipment. It is thus necessary to reduce time and man power through an effective method of preform design. In this paper, the equi-potential lines designed in the electric field are introduced to find an appropriate preform shape. The equi-potential lines generated between two conductors of different voltages show similar trends for the minimum work paths between the undeformed shape and the deformed shape. Based on this similarity, the equi-potential lines obtained by the arrangement of the initial and final shapes are utilized for the design of the preform, and then the artificial neural network is used to find the range of initial volume and potential value of the electric field. (C) 2002 Elsevier Science Ltd. All rights reserved.