Six methods of evaluating the degree to which multitrait-multimethod matrices meet the Campbell-Fiske criteria (1959) for construct validity are illustrated and evaluated. It is concluded that the path analytic method (Werts & Linn, 1970) provides the most detailed information regarding individual traits and methods as well as permitting the evaluation of alternate models of the data which involve rank reduction in methods, traits, or both. The analysis of variance technique (Kavanagh, et al, 1971) provides summary data based on the full matrix which can be useful for inter-matrix comparisons but does not assess the conformity of individual traits or methods to Campbell-Fiske criteria. The various factor analytically based approaches (Golding & Seidman, 1974; Jackson, 1975; Tucker, 1966) are most useful when one is interested in exploring the factor structure of a set of data across data collection methods.