Objective: Ultrasound is used increasingly in swallowing research because it can be a non-invasive, repeatable and cost effective measure of swallowing dynamics and rehabilitative effects. However unstable head position and transducer movement while imaging may result in measurement error that is substantive enough to bias research results. This study investigated the variation in measures of hyoid displacement and submental muscle size within and across three sessions using two methods of acquisition. Methods: Twenty-four healthy participants over the age of 50 attended three sessions. In each session, five dynamic video clips of hyoid movement and five static images of submental muscles were imaged in 2D ultrasound using two acquisition methods. One method involved manual hand-held stabilisation of the transducer and the other method involved a custom-designed stand for stabilisation of the transducer and participant. Hyoid displacement was measured as a percent change between measures made at rest and at maximal excursion. Additionally, cross sectional area (CSA) measurements were made of the paired geniohyoid and the left and right anterior belly of digastric (ABD) muscles at rest. Results: Out of 720 possible measures of hyoid displacement, 675 measures were analysed. There were no significant order effects of session or trial with changes that were <1% and no greater 4.5% for session and no greater than 1.5% for trial from estimated baseline measures. There was a significant effect of method (p<.01), with a systematic decrease in stand measures that were <9.5% and no greater than 16% from estimated baseline measures. Variance was larger across sessions than within sessions. The stand condition was more variable than the hand-held condition for measures of hyoid displacement. Out of a total of 2160 possible measures of submental muscle size, 555 measures of geniohyoid and 1408 measures of ABD muscles were analysed. There were no significant order effects of session, trial or method in geniohyoid muscle measures. The estimated order effects of session were <3% and no greater than 13%, and trials were <0.4% and no greater than 0.7% from estimated baseline measures. There were no significant order effects of session and method in ABD muscle measures with order effects of session that were <2% and no greater than 5.5%. There were significant effects of trial (p<.01) in both ABD muscles with a small systematic increase that was <0.5% and no greater than 0.8%. Variance was larger across sessions than within sessions in all measures of submental muscle size. The stand condition was less variable than the hand-held condition for all measures of submental muscle size. Conclusion: The results from this study provide guidance to researchers who intend to use repeated measures from ultrasound imaging as an outcome measure in swallowing research. The large variability within and across participants in measures of hyoid displacement and geniohyoid muscle size may require further investigation. When the variations described in this current study are considered in the measures of the ABD muscle size then it can be a valuable measure of rehabilitative techniques.