Background:Stimulus-evoked electromyographic changes can be recorded from the extraocular muscles. These short-latency negative-polarity evoked myogenic potentials are called ocular vestibular evoked myogenic potentials (oVEMPs). To date there has not yet been a large-scale study examining the effects of age on the amplitude, latency, threshold, and interaural differences of the oVEMP to air-conducted stimuli. Further, before the oVEMP can become a useful clinical tool, the test‐retest reliability of the response must be established. The oVEMP response, once more completely understood, may provide diagnostic information that is complementary to the cervical vestibular evoked myogenic potential (cVEMP; i.e., sternocleidomastoid muscle).Purpose:To describe the normal characteristics of oVEMP in a cohort of age-stratified subjects, to assess the test‐retest reliability of the oVEMP, and to determine if reference contamination occurs using a common recommended infraorbital reference electrode derivation.Research Design:A prospective, descriptive study design was used for an investigation with a threefold purpose in which oVEMP recordings were made from the extraocular muscles (e.g., inferior oblique muscle).Study Sample:Fifty otologically and neurologically normal adults and children served as subjects. Subjects ranged in age from 8 to 88 yr.Data Collection and Analysis:In Investigation 1, oVEMPs were recorded from the ipsilateral and contralateral inferior oblique muscles for all subjects. The stimulus was a 95 dB nHL 500 Hz tone burst. Next, oVEMP thresholds were obtained. Amplitude, latency, and thresholds were tabulated, and descriptive statistics were used to calculate normative values. Age-related differences in oVEMP component latencies, amplitudes, interaural amplitude asymmetries (IAAs), and thresholds were determined using an analysis of variance. In Investigation 2, oVEMPs were recorded twice in 10 subjects, once (test) and once approximately 10 weeks later (retest). Test‐retest reliability for the oVEMP peak-to-peak amplitude, n1 latency, p1 latency, n1 threshold, and IAA were assessed with intraclass correlation coefficients (ICCs) calculated using a two-way random-effects, absolute-agreement model. In Investigation 3, a four-channel oVEMP recording was conducted in 10 subjects. Both observational methods and paired-sample t-tests were used to evaluate the effect that reference electrode location had on the oVEMP.Results:oVEMP responses were present bilaterally in 90% of our subjects. The upper limit of oVEMP amplitude asymmetry, defined as the mean plus two standard deviations, was 34% (mean = 14%, SD 10), and the mean n1 latency was 12.5 (SD 1.0) msec. The amplitude of the response significantly decreased and the threshold significantly increased with increasing age, with the greatest age effects occurring in subjects 50 yr and older. Test‐retest reliability was acceptable (ICCs for the measurement variables ranged from .53 to .87). Using conventional recommended recording techniques, evidence of reference contamination occurred for all subjects, resulting in a mean amplitude reduction of 30% (range = 18%‐43%).Conclusions:Age results in systematic changes in oVEMP measurement parameters. The test‐retest reliability is acceptable, and reference contamination averaging 30% is guaranteed using a second infraorbital electrode as the inverting input (i.e., reference electrode) for bipolar recordings. The oVEMP can be used as a complementary diagnostic tool to the cVEMP in evaluating subjects with suspected peripheral vestibular disorders.