Objective: Accurate and economic measurement of energy expenditure (EE) is necessary to assess the effects of pediatric obesity therapy and its various modes of exercise. Flex heart rate (HR) monitoring, increasingly used in field studies, is a non-invasive, inexpensive method to predict EE from 14R. This study validated the flex HR method against indirect calorimetry in children attending an in-hospital 4-week obesity therapy. Methods: According to the flex HR method, HR-EE relationships were obtained in 12 obese children without comorbidities (12.9 +/- 1.8 years, BMI 30.6 +/- 3.9 kg/m(2), VO2max 29.8 +/- 4.5 ml/(kg*min)). Individual calibration and flex HR point definition were based on simultaneous recordings of VO2 and HR during three resting modes and a modified Bruce treadmill protocol. Five characteristic exercise therapy programs were selected for field-test validation. There, EE was assessed by indirect calorimetry (EEIndKal) and compared to the flex HR-based EE estimate (EEHF-Flex). Results: Mean differences between EEHF-Flex and EEIndKal for a 6-minute running test, ball games, cycle ergometry (65 W) and strength/stability circuit were +3.6 +/- 15.4%, +9.4 +/- 16.1%, +14.7 +/- 20.1% and +28.1 +/- 27.8%, respectively. Pearson's correlation coefficients ranged from r=0.92 (running, p < 0.001) to r=0.76 (strength/stability circuit, p=0.01). Discussion: The validity of EEHF-Flex in obese children depends largely upon exercise mode. If calibration and field test activity correspond, the flex HR method provides a satisfactory estimate of group EE. The method overestimates actual EE in intermittent or strength-related exercises. It is not suitable for individual estimates.