During clinical radiofrequency catheter ablation a wide range of delivered power may be necessary to achieve success despite an apparently stable catheter position on fluoroscopy. The purpose of this study was to use intracardiac echocardiography to characterize the relation between catheter tip–tissue contact and the efficiency of heating during applications of radiofrequency energy in vivo and to determine whether intracardiac echocardiography could be used prospectively to improve tissue contact. A closed-loop temperature feedback control system was used during radiofrequency applications at five anatomic regions in the right atrium of 15 anesthetized dogs to ensure achievement of a predetermined temperature (70° C) at the catheter tip thermistor by automatic adjustment of delivered power (maximum 100 W). The efficiency-of-heating index was defined as the ratio of steady-state temperature (degrees Celsius) to power (watts). Two-dimensional intracardiac echocardiography was used to evaluate movement of the catheter tip relative to the endocardium. Perpendicular contact was scored as good, average, or poor and lateral catheter sliding as <2, 2 to 5, or >5 mm. Two groups of animals were included: group 1, in which tissue contact was guided by fluoroscopic and electrographic criteria for stability of contact, with intracardiac echocardiography used simply to observe the application; and group 2, in which tissue contact was guided by intracardiac echocardiography. Of 66 applications, 18 (27.3%) had poor perpendicular contact on echocardiography, and 12 (18.2%) demonstrated lateral sliding of >5 mm even though they had been considered to have good tissue contact by fluoroscopic and electrographic criteria. Perpendicular catheter contact and anatomic location were shown to be independently related to the efficiency-of-heating index. Applications with good perpendicular contact had a significantly higher efficiency-of-heating index and a significantly greater lesion size than those with average or poor contact. The percentage of applications having good perpendicular tissue contact and the lesion size were significantly greater when tissue contact was guided by intracardiac echocardiography compared with fluoroscopic and electrographic guidance. This study demonstrates that variations in catheter tip–tissue contact account for differences in the efficiency of tissue heating, independently of the anatomic site of the application. Poor tissue contact was observed by intracardiac echocardiography and confirmed by indexes of tissue heating in approximately one third of radiofrequency applications despite a fluoroscopic appearance and electrographic morphologic appearance suggestive of good tissue contact. There was a significant correlation between echocardiographic evaluation of tissue contact, parameters of tissue heating (efficiency-of-heating index), and lesion size. In addition, intracardiac echocardiography could be used prospectively to improve the percentage of good contact applications and increase the lesion size. (Am Heart J 1997;133:8-18.)