Physiological Cost Index (PCI) of walking has been widely used to predict oxygen consumption in healthy subjects or patients. The purposes of this study were to evaluate the predictability of physiological cost index of upper extremity exercise with a sander for the amount of exercise and cardiac function and to find out most efficient speed of exercise. Upper extremity exercise was conducted in 67 healthy children (age 4 -12) for 2 minutes with a sander in two different speeds, such as self selected comfortable and self selected fast speeds. Speed of sander was calculated, and heart rate were measured before and immediately after the exercise. Physiological Cost Index (PCI) of upper extremity exercise was calculated for statistical analysis. The result were as follows; 1.　Comfortable and fast speeds of the sander went up according to the increase of age, sander PCI during exercise by it, decreased as the age increased with significant difference in each group (p=0.00). The heart rate change during exercising by it, there was no significant difference. 2.　The heart rate change during the exercise by sander increased as the speed of sander increased. Linear regression equation between the heart rate change during the exercise by sander and the speed of sander were 'Y (the heart rate change by sander) = 0.332X (the speed of sander) + 12.731' (R²= 0.154). Stepwise regression showed the speed of sander affected the change of heart rate positively, and the age affected negatively (R²= 0.242, p = 0.00). 3.　The PCI during the exercise by sander decreased as the speed of sander increased. Linear regression equation between the PCI during the exercise by the sander and the speed of sander were 'Y (the PCI by sander) = -0.00495X (the speed of sander) + 2.091'(R²= 0.267), and statistic significant was recognized (p = 0.00). 4.　The speed of exercise by sander it increased as the age increased. Linear regression equation between the speed of the sander and the age were 'Y (the speed of sander) = 2.481X (the age) -0.893(R²= 0.452), and statistic significant was recognized(p = 0.00). 5.　The reliability analysis for the method of examining each PCI, showed high reliability coefficients and the value of alpha was 0.9337. In conclusion, PCI of the upper extremity exercise with a sander could be thought to predict the energy consumption from the speed of sander as there was a linear relationship between the speed of sander exercise and heart rate change with exercise. PCI of the upper extremity exercise with the self selected fast speed was most efficient. Further study with oxygen consumption analysis is recommended to apply PCI in upper extremities in patients with different pathologies in motion.