Active virtual games (AVG) may make possible gross motor skill development, depending on their fidelity. This study compared the movement patterns of nineteen 10-12yr old children, whilst playing table tennis on three AVG consoles (Nintendo Wii, Xbox Kinect, Sony Move) and as a real world task. Wrist and elbow joint angles and hand path distance and speed were captured. Children playing real table tennis had significantly smaller (e.g. Wrist Angle Forehand Real-Kinect: Mean Difference (MD): -18.2°, 95% Confidence Interval (CI): -26.15 to -10.26) and slower (e.g. Average Speed Forehand Real-Kinect: MD: -1.98ms-1, 95% CI: -2.35 to -1.61) movements than while using all three AVGs. Hand path distance was smaller in forehand and backhand strokes (e.g. Kinect-Wii: MD: 0.46m, 95% CI: 0.13-0.79) during playing with Kinect than Move and Wii. The movement patterns when playing real and virtual table tennis were different and this could hinder the development of real world gross motor skills. Several elements, including display, input and task characteristics, might have added to the differences in movement patterns observed. Understanding the interface components for AVGs may assist development of higher fidelity games to potentially enhance the development of gross motor skill and thus participation in PA.