This research attempted to determine which kinds of free-choice motor activity improve balance control. One hundred and fifty children from three kindergarten classes participated in the research. They were provided with an environment that enabled them to choose motor activity freely for 180 minutes each day, 90 minutes outdoors and 90 minutes indoors, five days a week for 32 weeks. The difference between the three kindergartens consisted of the number of facilities intended for balance exercise and their location indoors or outdoors. The findings showed that children attending kindergartens in which the balance facilities were located both indoors and outdoors had better achievements than those attending kindergartens having balance facilities only indoors or only outdoors. A regression analysis pointed to only social types of free-choice motor activity having a significant impact on improvement of balance control. Further research is suggested to determine the optimal social and physical environments that enhance balance control. (author abstract)