Abstract: Neurolinguistic programming's hypothesized eye-movements were measured independently from videotapes of 30 subjects, aged 15 to 76 yr., who were asked to recall visual pictures, recorded audio sounds, and textural objects. Chi square analysis indicated that subjects' responses were significantly different from those predicted. When chi square comparisons were weighted by number of eye positions assigned to each modality (3 visual, 3 auditory, 1 kinesthetic), subjects' responses did not differ significantly from the expected pattern. These data indicate that the eye-movement hypothesis may represent randomly occurring rather than sensory-modality-related positions.