he purpose of the study was to investigate the relationship between body mass index (BMI) and spatio-temporal gait characteristics of overweight/obese and non-obese school-aged children (12-14 years) at two different walking speeds. Eighty-four overweight/obese (n=28; age: 13.96 ± 0.79 yrs; mass: 74.8 ± 18.21 kg; height: 159.2 ± 7.1 cm and BMI: 29.28 ± 5.64 kg/m²) and non-obese students (n=56; age: 13.72 ± 0.79 yrs; mass: 51.7 ± 10.2 kg; height: 157.8 ± 8.3 cm and BMI: 20.69 ± 2.74 kg/m²) with no present injuries were recruited. Participants were instructed to walk across an electronic walkway in each of two experimental conditions: a self-selected comfortable walking speed and a “walk more quickly” speed. Dependent variables of interest were cadence, gait velocity, step length (left and right), base of support (left and right) and percent double support (left and right). Independent t-tests reported a significant difference for BMI between groups (p < 0.000). Results for 2 (group) x 2 (speed) mixed model ANOVAs identified no significant interactions, while walking speed produced significantly different velocity, cadence, step length, and percent double support characteristics. Bilateral double support percent and bilateral base of support were significantly different between groups. It was observed that the noncontributory mass (additional excess fat) possessed by overweight/obese children may contribute to biomechanical inefficiency of movement and impaired stability. Many growing children, more commonly obese children, display considerable disruption to normal spatio-temporal gait characteristics when walking at a slower or more quickly and normal walking pace. The obese group showed pronounced alteration in gait for both base of support as well as double support (% gait cycle). These changes have been interpreted as representing underlying instability in obese children, with a normal, comfortable walking speed and longer periods of double support and base of support thought to assist with the maintenance of dynamic balance when performing everyday movement tasks. An exploratory multiple regression analysis was performed to predict BMI as the dependent variable from gait-related variables identified (n=17) for each walking speed. Results identified double support time was a primary predictor of BMI in 12-14 year old children. The unreliability of gait patterning observed in obese children is related to body composition and is affected by speed of walking. The evaluation of gait may also provide an indication of potential problems with the persistence of obesity.