A new lead design was tested that combined a small microporous steroid-eluting electrode with an insulated, exposed helix for active fixation. This lead (model 5078, Medtronic, Inc., group I, n = 10) was compared to a conventional model (model Y 60 BP, Biotronik) with a larger surface of polished platinum-iridium, equipped with a fixed, noninsulated screw but without steroid elution (group II, n = 10). The two lead models were studied in the atrial position of dual chamber pacing systems, which all had a tined ventricular lead (model 5024, Medtronic, Inc.), with essentially the same steroid-eluting tip as the new active fixation lead design. Sensing and pacing data were recorded acutely and during 1 year of follow-up, via the telemetry of a Relay pulse generator (Intermedics, Inc.). Intraoperatively, unfiltered atrial electrogram amplitudes did not differ between groups (group I: 7.12 +/- 2.56 mV vs group II: 6.42 +/- 1.87 mV; P > 0.05), nor did sensing thresholds 1 year after implantation (group I: 5.33 +/- 1.70 mV vs group II: 4.26 +/- 1.40 mV; P > 0.05). Atrial pacing thresholds as measured during surgery at a pulse width of 0.5 msec were lower in group I (0.49 +/- 0.15 V) than in group II (0.68 +/- 0.19 V; P < 0.05). From day 5 through day 360 of follow-up, the difference in atrial pacing thresholds was highly significant (P < 0.01), with a smaller peaking of early thresholds and a much lower scattering of data for the steroid screw-in leads than for controls.(ABSTRACT TRUNCATED AT 250 WORDS)