There is mounting evidence that human muscles have discrete partitions. In an effort to accurately assess how the control to each of the three heads (designated as A-C) of the lateral gastrocnemius (LG) muscle might be organized, we performed exploratory studies in five subjects each of whom was provided 30 random angular and linear perturbations while standing on a balance platform in each of three experiments. The responses from each LG head were recorded with strategically-placed fine-wire electrodes and, after correction for variation in distance to each recording site, latency to activation and the sequence of recruitment for each LG head were determined. Within individual subjects, the same LG head was first recruited. The C-head was recruited first during linear perturbations between subjects and across sessions, but the overall recruitment pattern was different between subjects. The fact that a consistent, selective response was seen during dynamic linear but not angular perturbations suggests that a more consistent strategy of inter-partition response may be engaged for the former perturbation. This behaviour may be caused by use of the LG C-head to assist the medial gastrocnemius to control ankle plantar flexion and pronation during the shearing forces generated with linear translations. Further exploration into the relationship between selective activation of muscle partitions in the presence of defined movement conditions appears justified.