Results of 40 experimental trials designed to enhance communication skills of patients or health-care providers are summarized. Calculation of effect sizes (ES) offered a method for comparing the impact of these interventions on communicative behaviors. Communicative behaviors were categorized into interpersonal processes and interviewing skills. For patient-targeted interventions that examined interpersonal process variables (n = 10), modeling and skill-practice interventions exhibited the largest ES values. For provider-targeted interventions (n = 30), 21 assessed interviewing skills, 5 examined interpersonal processes and 4 examined both categories. Effect sizes were moderate to large for interviewing skills interventions targeted at medical students. For other provider groups, the majority of studies failed to include the necessary data to calculate ES. The dearth of studies examining interpersonal processes prevented drawing any conclusive statements about their relative efficacy. On the basis of these findings, current methodological concerns are described, recommendations for future research presented and practical implications discussed. © 1991.