Abstract: Tested the neurolinguistic programming hypothesis that most people have a primary representational system for dealing with the world. 50 undergraduates were evaluated for sensory modality preferences in 3 ways: (1) They chose among written descriptions using wither visual, auditory, or kinesthetic wording; (2) their eye movements were recorded during an interview; and (3) their verbal responses were scored for sensory predicates. Results did not support neurolinguistic programming theory in that preferences of one modality on one measure did not relate to the same modality on the other measures as would be expected if primary representational systems were characteristic of the sample. It is concluded that, on the basis of both the results of the present study and mixed results obtained in other studies, more empirical support is needed before the positive therapeutic claims of neurolinguistic programming proponents can be accepted.