The activation of skeletal muscle during voluntary isometric contraction has been assessed by measuring the increase in force caused by a superimposed maximal shock to the motor nerve (the twitch-interpolation technique). When the muscle is held isometric, the increase in force with stimulation (superimposed twitch force) decreases with increasing voluntary force, and a line fit through the data can be extrapolated to maximal voluntary force at the zero twitch force axis. In a previous paper we questioned the applicability of this technique in situations where a high series compliance allows the muscle to shorten during the superimposed twitch. To explore effects of series compliance, we measured force of the adductor pollicis during voluntary isometric contractions with noncompliant and compliant loading devices. With the compliant loading device, superimposed twitch force was systematically less than with the noncompliant device, and the plot of superimposed twitch force vs. voluntary force was often concave upward, preventing easy extrapolation to maximal voluntary force. These findings are consistent with force-velocity characteristics of muscle and suggest that twitch-interpolation data must be interpreted with caution when the muscle is not held isometric during the superimposed twitch.