The redundancy statistic (Rd) is discussed in relation to canonical correlation analysis. The index is a measure of the variance of one set of variables predicted from the linear combination of the other set of variables. A small data set (N=6) from the work of D. Clark (1975) was analyzed using SPSS-X. Two sets of two variables each were included. The variables of one set were considered criterion variables, and the variables of the second set were considered predictors. The CANBAK computer program was used to obtain squared structure coefficients, communality coefficients, and variate adequacy coefficients. The results suggest that the Rd was not truly multivariate and may have more value in the context of redundancy analysis than in canonical correlation. Rd is more a measure of prediction than association. It is an estimate of the shared variance of linear composites, not the variables themselves. The Rd is analogous to the squared multiple R in multiple regression. Redundancy coefficients computed in a canonical analysis may have somewhat limited value. Four tables present illustrative data. A 29-item list of references is included. (SLD)