Introduction: Post-surgical breast radiation therapy (RT) is effective in reducing the recurrence of cancer but is known to increase long-term risk of cardiovascular events in patients. Deep inspiration breath hold (DIBH) is one method to mitigate these risks of RT by spatially separating radiation sensitive targets from the treatment field. Calypso (Varian Medical Systems, Palo Alto, CA) is an electromagnetic (EM) tracking system that allows localization and monitoring of transponders during RT. Purpose: The current research was conducted with the objectives of (a) to confirm the known accuracy of Calypso surface beacons while tracking a phantom in motion and to quantify effects of eddy currents in the presence of metallic objects, (b) to validate the use of the Calypso surface beacon as a method of tracking chest wall motion in volunteers with under 5 mm total uncertainty, and (c) to determine if there exists a measurable difference in reproducibility of DIBH between coaching methods. Methods: A phantom study was conducted to investigate the electromagnetic interference via metallic objects placed within the electromagnetic field used by the Calypso array. Different configurations of conductive objects were used to test eddy current interference effects. A volunteer study with five healthy volunteers was conducted with 10 - 20 breath holds per session, each with 20 second duration while surface beacon position was monitored. Results: The total of positional uncertainty in the phantom study was 0.062 cm, 0.059 cm, and 0.061 cm for the lateral, anteroposterior (AP), and superior-inferior (SI) dimensions respectively. The uncertainty introduced by the presence of conductive objects was below 0.1 mm across all tested scenarios and thus considered clinically insignificant. Volunteer intrafraction error was dominated by inconsistency in achieving target DIBH levels. No statistically significant difference in breath hold consistency or variance was found across the volunteers between coaching methods. Mean intrafraction errors in AP and SI beacon tracking were 0.11+/- 0.06 cm and 0.10 +/- 0.05 cm, respectively. The overall uncertainty was below the targeted 3 mm standard error in all volunteer cases.