This study assesses the clinical and electrode complications associated with the use of the epicardial screw-on electrode technique for permanent ventricular pacing. The records of 114 consecutive patients who received screw-on epicardial electrodes were analyzed in-hospital and at one year following implantation. We compared two techniques of implanting the screw-on electrode: 72 transmediastinal (TM) and 42 subxiphoid (SX) procedures. Baseline clinical inequalities were not apparent between the SX and TM patients. There was one operative death in the SX series and none among the TM patients. Total hospital mortality was 4% in the SX group and 2% in the TM group. In-hospital morbidity in the SX series (13/72-18%) was less than the TM group (14/42-33%), largely due to pleural entry and necessity for more frequent chest tube placement in TM patients. Electrode complications were low in both groups (3% in-hospital and 3% late; total 6/114, or 5%). Morbidity and mortality of these approaches to direct myocardial pacing remain noticeably higher than reported results of some transvenous series. Excellent electrode stability can be attained using the screw-on electrode, however, and the SX approach offers lower hospital morbidity than the TM method.