Vocal hyperfunction is a description of abnormal patterns of vocal behavior that may lead to many common voice disorders. Previous studies demonstrated that disorders associated with hyperfunction could be detected in patients by measuring aerodynamic and acoustic parameters from recordings of a single sustained vowel using a Rothenberg mask setup. Although ambulatory systems have shown the best potential for unobtrusive long-term monitoring of vocal function, their ability to differentiate hyperfunctional from normal patterns of vocal behavior has not been assessed. This study provides an initial quantitative evaluation of the capabilities of a neck surface acceleration signal to objectively detect abnormal vocal behaviors associated with hyperfunctionally-related disorders. The goal is to verify if such detection is possible using a neck accelerometer signal rather than an airflow mask and incorporate vocal gestures from multiple vowels and running speech. An impedance-based inverse filtering algorithm is used to estimate aerodynamic parameters from the neck-surface acceleration signal. The results obtained when contrasting five patients with vocal nodules to five paired normal subjects indicate that the accelerometer-based assessment offers comparable discrimination capabilities as those from the aerodynamic recordings. The results also provide a first indication that this discrimination is possible with an expanded sample that includes other sustained vowels and running speech.