Postural stability is reduced with age or neurological disorder and improved with certain types of training (e.g., Tai-chi) as evaluated by various measures using center of pressure (COP) during quiet standing. However, the implications of specific measures with respect to the underlying control mechanisms have remained unclear. PURPOSE: To investigate the influence of variation in the control system on the COP measures during quiet standing, using mathematical simulations. METHODS: A feedback model that represented the human body and the control system during quiet standing was built using an inverted pendulum, a proportional and derivative (PD) controller with a time delay (the neural controller) and another PD controller without a time delay (the mechanical controller). The angle of the pendulum was fed back to the controllers, and the output of the neural controller (corresponding to the motor command) was input into the neuro-muscular torque generator. The sum of the output of the neuro-muscular torque generator and of the mechanical controller represented the ankle torque. A driving noise was added to the motor command to simulate realistic body sway conditions. The gain combinations that robustly stabilized the entire system were used to simulate the COP fluctuations. The gains of the controllers and the noise amount were treated as variables. The statistical relationship between each measure and the controller gains or noise was assessed using the partial correlation analysis. Sixty thousand and five hundred combinations of the variables were tested in total. RESULTS: All gains correlated negatively with the time domain measures (r=-0.372 to-0.659) except the mean velocity, which correlated positively with each gain. The neural derivative gain correlated positively with the frequency domain measures (r =0.964 to 0.973). The proportional gains correlated with Hurst exponent positively for the short term region (r= 0.837 to 0.892) and negatively for the long term region (r=-0.828 to-0.883). The amount of noise correlated positively with time domain measures (r= 0.901 to 0.988) but not with frequency measures. CONCLUSIONS: Each COP measure is associated with the variation in the control system. As such, the COP measures have a potential for investigating or characterizing the control system of balance.