A series of in vivo trials were carried out in order to verify whether the electromagnetic field radiated by GSM (Groupe Systemes Mobiles) mobile cellular phones might affect implanted pacemakers. Two European GSM phones of 2-watt power were tested and trials conducted on 101 pacemaker implanted outpatients attending day hospital for routine check-up, who volunteered for trials. Forty-three pacemaker models from 11 manufacturers were tested in all. When the sensing threshold of the pacemakers was set at a minimum and the antenna of the phone was in direct contact with the patient's chest, interference was detected for 26 implanted pacemakers. Specifically, pulse inhibition in 10 of 101 cases, ventricular triggering in 9 of 46 DDD-VDD pacemakers, and asynchronous pacing in 4 of 52 devices. Pulse inhibition was also observed combined with asynchronous pacing in 1 of 52 cases and with ventricular triggering in 2 of 46 cases. Minimum effect duration was ca. 3 seconds but in 6 cases effects continued as long as the interfering GSM signal was on. No permanent malfunctioning or changes in the programmed parameters were detected. Whenever interference was detected, trials were repeated to determine the maximum sensing threshold at which interference persisted (with the antenna in contact with the skin over the pacemaker). Then maximum distance between antenna and pacemaker at which interference occurred was determined at pacemaker maximum and minimum sensing threshold. Under our experimental conditions electromagnetic interference effects were detected at a maximum distance of 10 cm with the pacemaker programmed at its minimum sensing threshold.(ABSTRACT TRUNCATED AT 250 WORDS)