Patient-centered care has received significant attention and is an integral component of high-quality healthcare. While it is often assumed that most prefer a patient-centered role orientation, such preferences exist along a continuum with some patients preferring a more provider-centered role. The present study examines patient preference data from a randomized clinical trial designed to test the efficacy of a patient activation intervention to promote thiazide prescribing for veteran patients with uncontrolled hypertension. Patient preferences for involvement in healthcare were assessed using the 9-item Sharing subscale of the Patient-Practitioner Orientation Scale (PPOS). The primary aim was to examine differences in discussion of thiazide use in the clinical encounter by those scoring high versus low on the PPOS. Five hundred ninety-five veteran patients were randomized to either one of three intervention groups or a usual care control group. The adjusted odds ratios (OR) for the three intervention groups relative to the control group indicated that thiazide discussion increased as a function of intervention intensity across both high and low PPOS groups. ORs for the most intensive intervention group were 3.72 (95% CI = 1.61-8.65, p < .01) for high PPOS patients and 6.71 (95% CI = 2.59-10.67, p < .001) for low PPOS patients. Results suggest that this patient activation intervention is effective for veteran patients representing a range of preferred involvement. Consideration of such preferences may be useful in tailoring future interventions in the healthcare context.Published by Elsevier Ltd.