Background Using physical examination to make the diagnosis of shoulder instability can be difficult, because typical examination maneuvers are qualitative, difficult to standardize, and not reproducible. Measuring shoulder translation is especially difficult, which is a particular problem, because measuring it inaccurately may result in improper treatment of instability. Questions/purposes The objective of this study was to use a magnetic motion tracking system to quantify the effects of external rotation of the abducted shoulder on a simulated simple translation test in healthy subjects. Specifically, we hypothesized that (1) increasing external rotation of the abducted shoulder would result in decreasing translation; (2) intraobserver repeatability would be less than 2 mm at all external rotation positions; and (3) mean side-to-side differences would be less than 2 mm at all external rotation positions. Methods The intraobserver repeatability and side-to-side differences of AP translation were quantified with a noninvasive magnetic motion tracking system and automated data analysis routine in nine healthy subjects at four positions of external rotation with the arm abducted. A shoulder positioning apparatus was used to maintain the desired arm position.