This study assessed the exercise tolerance and the cardiorespiratory responses to a training program by the six-minute walk test (6'WT) in children with congenital heart disease (CHD). Seventeen cardiac and 14 healthy children performed maximal cardiopulmonary exercise test (CPET) and 6'WT. Reliability of 6'WT was assessed in all subjects (test-retest) by Bland-Altman plots. Cardiac subjects were randomly divided in training (T-CHD) and control groups (C-CHD). T-CHD underwent an individualized training exercise at the ventilatory threshold (VT) intensity during 12 weeks. We found that the 6'WT is a reliable and reproducible test. CHD children walked a lower distance than healthy children before training (472.5 +/- 18.1 vs. 548.8 +/- 7.7 m, respectively, p < 0.001). Likewise, power output, oxygen uptake (V.O (2)), and heart rate (HR) at the maximum and the VT levels, were significantly lower in patients (p < 0.001). After training, a significant improvement of walking distance (WD) was shown in T-CHD (529.6 +/- 15.3 vs. 467.7 +/- 17.1 m, p < 0.001). The power output, VO2, HR, and V.E increased slightly (6 to 10 %, p > 0.05) at peak exercise and significantly at ventilatory threshold level (p < 0.05) in T-CHD. Significant relationships between WD and VO2max as well as VO2 at VT were founded (p < 0.05). We concluded that the 6'WT is a useful and reliable tool in the assessment and follow-up of functional capacity during rehabilitation program in children with CHD.