Electromagnetic tracking devices provide a flexible, low cost solution for three-dimensional ultrasound (3-D US) imaging. They are, however, susceptible to interference. A commercial device (Ascension pcBIRD) was evaluated to assess the accuracy in locating the scan probe as part of a digital, freehand 3-D US imaging system aimed at vascular applications. The device was optimised by selecting a measurement rate and filter setting that minimised the mean deviation in repeated position and orientation measurements. Experimental evaluation of accuracy indicated that, overall, absolute errors were small: the RMS absolute error was 0.2 mm (range: -0.7 to 0.5 mm) for positional measurements over translations up to 90 mm, and 0.2 degrees (range: -0.8 to 0.9 degrees ) for rotational measurements up to 30 degrees. In the case of position measurements, the absolute errors were influenced by the location of the scanner relative to the scan volume. We conclude that the device tested provides an accuracy sufficient for use within a freehand 3-D US system for carotid artery imaging.