Cross-validation, an economical method for assessing whether sample results will generalize, is discussed in this paper. Cross-validation is an invariance technique that uses two subsets of the data sample to derive discriminant function coefficients. The two sets of coefficients are then used with each data subset to derive discriminant function scores. The scores are correlated to assess the stability of the discriminant function coefficients across samples. The closer the set of results are to each other, the greater the stability of the coefficients is across samples. One advantage of cross-validation is the ease of calculation and adaptability to other statistical procedures. A weakness is that the sample size must be large in order for division of data into data subsets to be meaningful. Six steps are outlined that demonstrate a double cross-validation invariance procedure in a discriminant analysis. The data set used consists of 64 subjects with two continuous predictor variables "x" and "y" and one criterion variable "Group" with four levels. It is noted that to evaluate the stability of discriminant function coefficients, the discriminant function scores themselves must be examined. Two tables are included, and an outline of the computer commands for simulating data in the first table is appended. (Author/RLC)