This monograph describes an approach to generating continuous multivariate distributions. Typically, the approach throughout is to use a transformation of one or two easily generated random variables to achieve a new distribution. A few parameters provide a mechanism for generating one of a family of possibilities. To help choose the correct parameter values, the reader is then presented with an array of three-dimensional and contour plots for the families.

A large number of distributions are described. These include the well known, the less well known, and the rare. The multivariate normal and multivariate uniforms are examples of the first type. The somewhat less well known or rare include the lognormal, logistic, Pareto, Burr, and Wishart, to name some of the remaining ones. This work is a serious contribution to the field and could be of use to those doing Monte-Carlo or discrete event simulations.