Factors affecting outcomes of medication-history interviewing by pharmacy students were studied. Data were obtained from fourth-year pharmacy students enrolled in a required course in fall 1984. Each student conducted a medication-history interview with one of two simulated patients who presented a predetermined history; interviews were videotaped from behind a one-way mirror. Students also completed an interviewing-orientation survey and a personal report of communication apprehension (PRCA). Trained raters evaluated the videotaped interviews using measures of interview skill and interview completeness. The simulated patients completed a patient-satisfaction form after each interview. Two path models were developed that were identical except that one had completeness and one had patient satisfaction as the dependent variable. Interview skill was the final factor in each model, preceded by variables representing the student's background and orientation factors, PRCA, and simulated-patient gender. Of 112 students conducting the interview, 107 (95.5%) and 95 (84.8%) completed the PRCA and orientation surveys, respectively. The models explained 36% and 27% of the variance in patient satisfaction and completeness, respectively. Shown in parentheses are the significant direct predictors of variables in the model of patient satisfaction: satisfaction (skill, prepharmacy grade point average [preGPA], people and health-care [PHC] orientation); skill (interviewing orientation, preGPA); interviewing orientation PHC orientation, preGPA, PRCA); and (PHC orientation (student gender). All effects were positive except for PRCA on interviewing orientation. For the model of completeness, direct predictors were as follows: completeness (skill, PHC orientation, student gender, simulated-patient gender); skill (interviewing orientation, preGPA); interviewing orientation (PRCA, preGPA, PHC orientation); and PHC orientation (student gender). All effects were positive except for PRCA on interviewing orientation and PHC orientation on completeness. Results suggest that one path model reflects the patient's assessment of interviewer competence in terms of satisfaction, and the other reflects the clinician-rater's assessment of interviewer competence in terms of interview completeness. The interviewing process positively influences both patient satisfaction and interview completeness.