To examine health promotion in a primary-care context, we studied perceived and actual communication in 271 consultations between general practitioners and patients in Oxford (England). Although health promotion is a term usually reserved for public-health or wellness programs, a health promotion perspective enriches the examination of communication in physician-patient interactions by emphasizing issues of empowerment, competence and control. Accordingly, we are interested in how communication during medical encounters can improve patients' abilities to exercise appropriate control over their health. A major factor in enabling patients to increase control over their health involves developing their competencies for making decisions and enacting behaviors that can lead to desired, and attainable, health outcomes. This report focuses on communication and decision making about prescription medications, since whether and how to use medications are among the most common and important decisions in which patients can participate. Five instruments were employed to collect data about physicians, patients and their consultations: a Video Analysis, which allowed assessment of actual communication behavior; a Patient Questionnaire designed to gauge perceptions of the encounter and collect demographic information; a Medical-Record Review, which provided information on utilization, diagnosis and treatment; a Telephone Interview, conducted 14 days after the consultation to obtain follow-up information (e.g. experience with the prescribed medication); and a Doctor Questionnaire that focused on attitudes toward consultations and patients. With respect to communication about prescription medications, physicians most frequently mentioned product name (78.2% of consultations) and instructions for use (86.7% of consultations). Patients were extremely passive, rarely offering their opinion or initiating discussion about any aspect of the treatment. We suggest that improving patients' decision-making competencies may require more discussion of benefits and risks, as well as discussion of patients' opinions about the prescribed medications and their abilities to follow through with the treatment plans. The research design proved useful in highlighting discrepancies between perceived and actual communication. Physicians tended to overestimate the extent to which they discussed patients' ability to follow the treatment plan, elicited patients' opinion about the prescribed medication and discussed risks of the medication. And, 24.3% of the patients left the consultation with an 'illusion of competence', a belief that important topics had been discussed when, in fact, they had not been mentioned at all. The pattern of results illustrates the complexity of health promotion in primary care, and underscores the importance of attending to both perceived and actual communication in medical encounters.