MANY of the more significant developments in computational technicoverlap basic contributions to statistical theory. It is difficult, indeed, to make clear cut distinctions between computational processes and statistical innovations. Some of the studies reviewed, therefore, must illustrate both theoretical and computational processes. During the past three years, the articles showing computational development deal primarily with the utilization of machines (particularly those of the International Business Machines Corporation) for mass processing of data, with methods for solving simultaneous equations, with methods of using analysis of variance, and with simplifications and extensions of factor methods. It must be recognized, however, that many contributions to mass methods of handling data developed or extended by the armed services may not be reviewed at present since these methods are classified as "restricted" or "confidential." There is good reason, however, to believe that in the next three years the computational methods developed in the armed services will become available to the research worker.