Market analyses generally consist of examining how consumption or ownership of a given product varies according to certain socio-demographic criteria such as sex, age, social class, area, town size, etc. Each criterion is considered separately to divide the population into a small number of groups. For marketing purposes, several consumption analyses according to different criteria are less useful than a single breakdown of the population into a limited number of groups so that the manufacturer of the product can evaluate the relative importance to attribute to each group in his marketing objectives. Obviously one can imagine numerous ways of breaking down the population into different groups, but segmentation will be more satisfactory when it leads to groups that are as clearly differentiated as possible. The object of this paper is to present a method of segmentation which enables one to obtain easily a result which corresponds to this objective. This in fact is not an original method but simply an adaptation of the technique proposed by Dr. BELSON for establishing matching samples.