Objective: To determine whether oral or nasal acoustic energy is primarily responsible for nonzero nasalance scores observed during the production of nonnasal sentences by individuals with normal speech. Method: Sixty adults with normal speech were asked to read the Zoo passage and produce three sustained vowels, (/i/, /a/ and /u/), with and without nares occlusion. Results: There was a significant decrease in nasalance scores between the unoccluded and occluded conditions for all four stimulus pairs. The mean decrease across conditions ranged from 8 (/u/) to 25 (/i/). In the unoccluded condition, the nasalance score was significantly greater for /i/ than for the other stimuli. Conclusions: The findings suggest that the majority of acoustic energy detected by the nasometer's nasal microphone during the production of nonnasal utterances is the result of sound transmission through the nose. The data obtained during this investigation, coupled with information available from other studies, suggest that this may be due to transpalatal transmission. If correct, such a conclusion would have clinical implications for patients with palatal clefts, since residual structural abnormalities and scar tissue in a repaired cleft palate may increase, dampen, or in some way alter transpalatal acoustic transmission. Thus, surgical normalization of velopharyngeal port control may not be sufficient to eliminate hypernasality in all patients.