BACKGROUND The purpose of this study was to evaluate electrode temperatures obtained using a radiofrequency ablation system that incorporates closed loop feedback control to achieve preset target electrode temperatures and to determine if closed loop temperature control results in a lower incidence of developing a coagulum. RESULTS Two hundred seventy patients underwent catheter ablation of atrioventricular nodal reentrant tachycardia, an accessory pathway, and/or the atrioventricular junction using an ablation system incorporating closed loop feedback control. Forty-five patients underwent catheter ablation in the power control mode in which power output was fixed, and 225 patients underwent catheter ablation in the temperature control mode. A coagulum occurred during 0.8% of radiofrequency applications in the temperature control mode versus 2.2% in the power control mode (P < .01). Electrode temperatures were within 10 degrees C of the targeted temperature during 35% of applications in the temperature control mode. Ability to achieve the targeted electrode temperature was related to the target, with radiofrequency energy applications at the atrioventricular junction resulting in the highest temperatures (70 +/- 12 degrees C) and those for ablation of the atrioventricular node the lowest (59 +/- 11 degrees C, P < .001), using a maximum of 50 W of power for both. Electrode temperatures were higher during ablation of left free wall and posteroseptal pathways than during ablation of right free wall and septal pathways. The mean and minimum temperatures associated with success were 64 +/- 12 degrees C and 44 degrees C, respectively. Overall, the electrode temperatures at successful and unsuccessful ablation sites did not differ (P > .05). CONCLUSIONS Temperature monitoring with closed loop control of power output facilitates radiofrequency catheter ablation procedures by minimizing the probability of developing a coagulum while ensuring maximum lesion formation.