This work explores the need for multiprocessing in the power industry. A suitable computer architecture for power system problems is presented. An efficient decomposition technique is developed to solve the problem parallely. Simulation of distributed processing using IBM 370 computer is considered. A microprocessor version of load flow program is developed and distributed load flow for on-line and off-line applications are studied. A novel algorithm for economic dispatch with and without transmission losses is presented. The algorithm utilizes a closed form expression for the calculation of the Lagrange multiplier thereby avoiding any iterative process in the calculation. Different nonlinear programming techniques are compared to the improved method. The algorithm presented is fast and appears to have good convergence properties. The improved algorithm is extended to distributed processing. Characteristics of a distributed microcomputer system and a multiple processor system are discussed. Intel 8080, 8085 and 8086 versions of distributed and multiprocessing methods are presented to solve power system problems. / Ph. D.