Physicians argue that patient preferences influence their test ordering and their potential for compliance with clinical practice guidelines (CPG). This study was conducted to evaluate patient satisfaction with clinical practice in emergency department (ED) settings using a validated and widely publicized set of CPGs. Patients presenting to 4 hospital EDs were eligible if they had sustained acute ankle or foot injuries. All sites were involved with the dissemination of radiography CPG, and use of radiography was determined by treating physicians. Telephone follow-up was attempted for all patients who did not receive ankle or foot radiography (Group 1). A random sample of 25% of patients who had a normal radiograph interpretation (Group 2) was also chosen for follow-up. Structured telephone interviews were administered and included information on post-encounter health care utilization, subsequent radiography, and patient satisfaction. A structured questionnaire was administered to all ED physicians (N = 60) to elicit their perspectives on the clinical practice guidelines. In Group 1, 342 (69%) of 494 non-radiographed patients were successfully contacted. In Group 2, 623 (77%) of 812 patients with normal ED radiographs, were successfully contacted. After ED discharge, 86 (25%) Group 1 and 191 (31%) Group 2 patients had visited another physician within 2 weeks of the initial ED encounter (P =.07). Subsequent ankle radiography was similar between the groups (38 [11%] in Group 1 vs. 59 [10%] in Group 2; P =.38). Patients appeared to be similarly highly satisfied with physician care (P =.58) and with discharge instructions (P =.12) in both groups. Overall, 76% of physicians supported the use of CPGs; however, 78% reported that patient expectations influenced their application of the Ottawa Ankle Rules. This study suggests that patients are equally satisfied with care, access additional health care services similarly and obtain the same percentage of radiographs irrespective of the initial ED ankle/foot radiograph ordering. These results may help physicians in re-evaluating their perceptions that patient expectation influence utilization and have important implications in guideline development.