Abstract: Bandler and Grinder (1975) have contended that individuals express themselves consistently in a language which is suggestive of a particular sensory modality called a "primary representational system" (PRS) and that by matching the PRS of another in one's own speech (i.e. "predicate matching") the relationship is facilitated. The present study investigated these assertions utilizing five trained counselors and forty undergraduate students in introductory psychology. The present study found little support for the stability of a person's PRS across different topics through the impact of predicate matching. Ss were met individuals and asked to speak extemporaneously into a tape recorder for one minute on each of three topics: a memorable vacation, an enjoyable meal, and a troublesome problem. Following the recording, Ss were played four cassettes (one for each PRS) prepared by the examiner to have the same topical content described above and to have a distinctive and highly saturated PRS. After each stimulus tape recording was presented, Ss completed either a simple rating measure involving a "like-dislike" dimension or they completed an Impact Message Inventory (Kiesler et. al., 1975). No significant consistency (Kendall Tau) was noted with regard to PRS material across topics. T-statistical evaluation for paired observations failed to confirm the experimental hypotheses that listeners would better "like" taped material sharing their own PRS (i.e., "congruence") nor were affiliation-related subscales of the IMI higher under conditions of congruence or mistrust-related subscales higher under conditions of incongruence. Results were discussed in light of other findings and in terms of some methodological shortcomings of the present investigation.