A 3-year study by three medical centers has revealed a 1-year electrode malfunction rate of 7.4%; most malfunctions occurred within the first 30 days. The incidence of unavoidable early malfunction (3.2%) fell within the 5% standards suggested by the committee report of the Inter-Society Committee on Heart Diseases. Incidences of obscure cause (3.2%) may be difficult to identify prospectively and may be, to a certain extent, unavoidable. The majority of the malfunctions (4.2%) showed specific clues that indicated that they were preventable. Successful repositioning was achieved on the first attempt in 80.6% of the cases with malfunction, and only 0.7% required ultimate myocardial electrode implantation. The principal clues to potentially unsatisfactory positioning included the presence of a large right ventricle with or without tricuspid insufficiency, current thresholds greater than 0.5 mA and ST-segment deviations on the intracardiac electrogram of less than 2 mV. Electrode malfunction may be more common with bipolar than with unipolar electrodes; but significant differences in the incidence of malfunction among different unipolar electrodes were observed. These data indicate that further developments in transvenous electrode design are warranted.