This paper is concerned with an iterative method for calculation of the capacity of a discrete, constant channel. Unlike the standard reference (Muroga 1953), in which Lagrange multipliers are used in the conventional form, the method described is based on techniques of mathematical programming. In addition, input signals of different duration are admitted. The algorithm proposed is very simple and has the advantage that it yields converging lower and uppper bounds for the capacity. © 1967 Academic Press Inc.