In a sustained movement task, such as human balance, continuous peripheral and serial ballistic intermittent control mechanisms might be integrated [1]. This implies that feedback mechanisms might not act continuously “closed loop” but can be intermittently “open”. Intermittent control limits the frequency bandwidth of the human operator and makes it refractory in nature. Classic experiments by Craik [2] revealed refractoriness in tracking responses to unpredicted, discrete step stimuli closely spaced in time. Recently, we employed a new method of analysis [3] and successfully identified refractoriness in the sustained visuo-manual control of unstable second-order systems [4]. It is, however, unclear whether the observation of refractoriness would generalize to a whole-body movement task like human balance. In the current study we designed an experimental setup (Fig 1 A) in which participants, standing on both feet, were strapped to a rigid structure that pivots at the ankle joint. This inverted pendulum model of human balance was actuated by a linear motor as a marginally stabilized second order system that run on a computer in real time. Participants controlled the system’s position through a myoelectric force signal derived from the combined EMG activity measurements of the tibialis anterior and the gastrocnemius medialis muscles. Visual feedback information about their human-attached-to-structure position was fed back to the participants by means of a sphere displayed on a large TV screen in front of them. The position of the pursuit tracking target was represented by a second sphere on the display (see Fig 1B for the step sequence). Our method of analysis showed that the goodness of fit of an appropriate order, zero delay ARMA model relating the control signal to the series of paired stimuli step disturbances could be improved by sequentially and individually adjusting the instant of each second step disturbance. This procedure resulted in a distribution of first- and second response times. Results show that the mean (n= 14) response time was significantly higher for the second than for the first stimulus (RT2: 311 ± 145 ms, RT1: 245 ± 145 ms, F(1, 13) = 15.9, p < .005). The interaction between Stimulus Number and Inter Step Interval, (F(7, 91) = 9.4, p < .0001) indicates that reducing the ISI had different effects on RT1 compared to RT2 (the two separate post-hoc tests both showed a significant ISI effect). Post-hoc test at each ISI level revealed that first and second response times were significantly different for ISIs up to 500 ms. These findings suggest that also for this whole-body balance task, the stimulus-response mechanism is refractory in nature; findings that might be better explained by an intermittent- than by a continuous control model.