The purpose of this study was to examine the reliability of the rating of perceived exertion (RPE), using the Borg 6-20 scale, at ventilatory threshold (VT) in children. Thirty children (19 male, 11 female) with a mean +/- SD age of 10.4 +/- 1.1 yrs performed two graded exercise tests for the assessment of VT and VO2max. RPE was recorded throughout each exercise test. There were no significant (p > 0.05) differences between the mean VO2max (50.9 +/- 6.0 vs 51.0 +/- 5.7 ml.kg-1.min-1; r = 0.95) or the mean VO2 at VT (36.2 +/- 4.4 vs 36.7 +/- 4.5 ml.kg-1.min-1; r = 0.87) between trials. The mean RPE at VT during trial 1 (12.4 +/- 2.7) was significantly higher than during trial 2 (11.4 +/- 3.3; p < 0.05). The test-retest reliability correlation for the RPE at VT was r = 0.78. The lower RPE in the second trial may have been due to a greater comfort and familiarity with the testing procedures. The reliability analysis indicates that the RPE at VT within a given child is fairly consistent from trial to trial. However, large inter-individual variability in the RPE at VT (range = 6 to 19) was noted and was not related to variations in the onset of VT. This raises some concern over the use of a given RPE value or range of RPE values in the regulation of exercise intensity for this age group. The test-retest reliability of VT and VO2max in this age group is similar to previous reports.