In conjunction with pharmacologic therapy, pacing has been proposed as a potential treatment to decrease symptoms in patients with moderate-to-severe congestive heart failure (CHF). Uncontrolled studies of pacing therapy for CHF dealing with different pacing sites, modes of pacing, and atrioventricular delays have reported mixed outcomes. The Pacing Therapies in Congestive Heart Failure (PATH-CHF) study is a single-blind, randomized, crossover, controlled trial designed to evaluate the effects of pacing on acute hemodynamic function and to assess chronic clinical benefit in patients with moderate-to-severe CHF. The effect of pacing on oxygen consumption at peak exercise and at anaerobic threshold during cardiopulmonary exercise tests, and on 6-minute walk distance, have been selected as primary endpoints of the study. Secondary endpoints of the trial were changes in New York Heart Association (NYHA) functional class, quality-of-life as assessed by the Minnesota Living with Heart Failure questionnaire, and hospitalization frequency. Finally, changes in ejection fraction, cardiac output, and filling pattern were assessed by echocardiography. The trial was planned to include 53 patients from 7 centers in Europe over a period of 3 years. The study was divided into 2 parts: acute testing and chronic follow-up. The acute study, performed during the pacemaker implantation, involved extensive testing using a custom-designed computer (FLEXSTIM) and a unique burst pacing protocol (FLEXSTIM protocol) to determine the best ventricular pacing sites and the most appropriate atrioventricular delays. The chronic phase consisted of a crossover study designed to test in each patient the best univentricular pacing site and biventricular pacing as assessed by the acute hemodynamic study. The study started with the first implant in 1995 and has, to date, included 42 patients. The study is expected to be completed by the end of 1998. The results of a first interim analysis showed trends toward improvement in all primary and secondary endpoints during the pacing periods compared with no pacing.