A method of estimating the likelihood ratio and the risk of an individual being affected with Martin Bell syndrome (MBS) from anthropometric measurements is described. The procedure is based on the discriminant functions (one for each sex), generated in our previous study in order to separate the individuals with MBS from the normal individuals. The procedure is illustrated by the examples of estimating the likelihood and the likelihood ratio in four individuals of either sex, belonging to MBS families, where the discriminant score value obtained from each individual is compared with the empirical (normalized) distribution of discriminant scores from the known MBS and normal subjects of a corresponding sex. The ways in which the risk of an individual being MBS is estimated in the general population or in members of the MBS families are indicated. The limitations of the discriminant diagnosis based on body measurements, as well as its particular applications in studies of the Martin-Bell syndrome, are discussed.