Congestive heart failure (CHF) patients with LBBB and QRS duration > 150 ms are considered the best candidates to biventricular pacing (Biv-P). However, patients with a narrow (120-150 ms) QRS may also benefit from Biv-P since true ventricular dyssynchrony may be underestimated by considering only QRS enlargement. From October 1999 to April 2002, 158 CHF patients (121 men, mean age 65 years, mean LVEF 0.29, mean QRS width 174 ms) underwent successful Biv-P implantation and were then followed for a mean time of 11.2 months. According to basal QRS duration, patients were divided in two groups, with wide QRS (> or = 150 ms, 128 patients, 81%) and with narrow QRS (< 150 ms, 30 patients, 19%). In the wide QRS group, LVEF improved from 29% to 39% (P < 0.0001), 6-minute walk test from 311 to 463 m (P < 0.0001), while NYHA Class III-IV patients decreased from 86% to 8% (P < 0.0001). In the narrow QRS group LVEF improved from 30% to 38% (P < 0.0001), 6-minute walk test from 370 to 506 m (P < 0.0001), and NYHA Class III-IV patients decreased from 60% to 0% (P < 0.0001). The data showed that in wide and narrow QRS patients, Biv-P significantly improved clinical parameters (NYHA class, 6-minute walk test, quality-of-life, and hospitalization rate) and main echocardiographic indicators. Furthermore, narrow QRS patients had a better survival rate, rapidly regained left ventricular function, and only a few patients remained in a higher NYHA class during follow-up. These patients should not be excluded "a priori" from cardiac resynchronization therapy.