OBJECTIVE This study investigated the association between neck length and vestibular evoked myogenic potential (VEMP) latencies in healthy children, adolescents, and adults to elucidate when VEMP latencies reach consistent levels. METHODS Findings of VEMP tests in 14 healthy children, seven healthy adolescents, and 14 healthy adults were analyzed for correlations with neck length, which was measured as the distance of a line dropping vertically from the mastoid tip to the horizontal plane passing through the clavicle. RESULTS All healthy children, adolescents, and adults exhibited present VEMP responses. Children, adolescents, and adults significantly differed in p13 latency, n23 latency, and p13-n23 interval. According to receiver operating characteristic curve analysis, the optimal cutoff values of p13 and n23 latencies between children and adults were 12.6 and 19.8 msec, respectively. Because the odds ratio of p13 latency was less than that of n23 latency, n23 latency was used to discriminate VEMP latencies between children and adults. Accordingly, a cutoff value of 15.3 cm for neck length was proposed as a criterion for predicting VEMP latency within the adult range. Consequently, a positive correlation between neck length and VEMP latency was observed when neck length was <15.3 cm, while above which level one need not account for neck length in evaluating VEMP latency. CONCLUSIONS The intra-subject variability of norms can be enhanced if the normative data for VEMP characteristic parameters take structural variance into account. This study suggests that the adult range of VEMP latencies can be anticipated if neck length is >15.3 cm.