The aim of this study was to determine the reproducibility of measuring the aerobic power and treadmill exercise economy in girls tested twice, 6 weeks apart We studied 61 normal-weight, prepubertal girls with a mean (± SD) age 7.3 ± 1.3 yrs and BMI 16.7 ± 1.5 kg/m2. Exercise economy (submax VO2) was determined while walking for 4 minutes at 2.5 mph with 0% grade. Aerobic power (VO2 max) was measured during a progressive, all-out, continuous treadmill test using standardized procedures and criteria. VO2, ventilation (V), respiratory quotient (RQ), and respiratory rate (RR) were measured using a Sensormedics metabolic monitor. Heart rate (HR) was measured using a Polar monitor. Most girls reached their physiologic limitas evidenced by the high proportion who reached criteria for VO2 maximum [RQ>1, HR>85th of age predicted maximum, and VO2 plateau]: 99% one criteria, 92% two criteria, and 70% all three criteria. Using paired t-tests, there were no significant differences between visits 1 and 2 in mean HR, RR, RQ, V, VO2 submax (421 vs 422 ml/min, respectively), and VO2 max (1036 vs 1049 ml/min, respectively). Coefficients of variation (CV) between visits 1 and 2 for submaximal tests were: HR=5.1%, RR=12.4%, RQ=7.2%, V=12.5%, and VO2= 12.4%. CVs for the maximum tests were: HRmax= 2.1%, RRmax=10.8%, RQmax=5.3%, Vmax=11.7%, and VO2 max=7.5%. These results indicate that treadmill testing for normal-weight, prepubertal girls is highly reproducible. Thus, one testing should give researchers an accurate measure of aerobic capacity and walking economy.