Based on Beisecker's earlier finding that patients over age 60 brought companions to their medical appointments more often than did patients aged 25-59, a study was conducted to examine the role of the companion. Eleven of 21 patients aged 60-85 brought companions. Companions were either spouses or adult children of the patients. Within this group, there was a significant difference between age and presence of a companion (mean age with companion was 72.36; mean age without companion was 65.60). There was no significant difference in length of interaction for patients with and without companions, indicating that the companion, by speaking, took time away from the patient. Tape-recordings of the 11 patients with companions were analyzed by coding comments in six interactive modes: doctor to patient, doctor to companion, patient to doctor, companion to doctor, patient to companion, and companion to patient. Interactions were divided into three segments: history, physical examination, and feedback. Results revealed that 15.65% of comments were directed from doctor to companion, whereas 22.15% were directed from companion to doctor, indicating that the companion responded or initiated comments when the doctor was not addressing him or her. Companion communication occurred in history and feedback segments with little activity during the physical examination. Companions played three roles: watchdog, significant other and surrogate patient. (Author/NB)