Head position and orientation are important variables during experiments conducted on the University of Wisconsin x-ray microbeam system, a nationally shared imaging facility for speech production research. Normally, the head is not restrained during experiments, in order to maximize the naturalness of speech performance, even though free head movements may have an adverse effect on the accuracy of kinematic data that are obtained. The basis for this effect is explained, and techniques for dealing with temporal variations in head position are described. Results obtained from a novel method for three-dimensional measurement of head position, applied to data sets recorded from six normal adult speakers during otherwise typical x-ray microbeam experiments, are summarized to convey some sense of the measurement error that can arise from deviations from a required head position. These measurements suggested that among these speakers, errors attributable to improper positioning of the head, and/or inadvertent head movements within or between trials, never exceeded 5%.