In this study, we compared measured maximal heart rate (HRmax) to two different HRmax prediction equations [22 - age and 208 - 0.7(age)] in 52 children ages 7-17 years. We determined the relationship of chronological age, maturational age, and resting HR to measured HRmax and assessed seated resting HR and HRmax during a graded exercise test. Maturational age was calculated as the maturity offset in years from the estimated age at peak height velocity. Measured HRmax was 201 +/- 10 bpm, whereas predicted HRmax ranged from 199 to 208 bpm. Measured HRmax and the predicted value from the 208 - 0.7(age) prediction were similar but lower (p < .05) than the 220 - age prediction. Absolute differences between measured and predicted HRmax were 8 +/- 5 and 10 +/- 8 bpm for the 208 - 0.7 (age) and 220 - age equations, respectively, and were greater than zero (p < .05). Regression equations using resting HR and maturity offset or chronological age significantly predicted HRmax, although the R2 < .30 and the standard error of estimation (8.2-8.5) limits the accuracy. The 208 - 0.7(age) equation can closely predict mean HRmax in children, but individual variation is still apparent.