Lumbar erector spinae (LES) contribute to spine postural and voluntary control. Transcranial magnetic stimulation (TMS) preferentially depolarizes different neural circuits depending on the direction of electrical currents evoked in the brain. Based on recent evidence, posteroanterior current (PA-TMS) and anteroposterior (AP-TMS) current would respectively depolarize neurons in the primary motor cortex (M1) and the premotor cortex. These regions may contribute differently to LES control. This study examined whether responses evoked by PA- and AP-TMS are different during the preparation and execution of LES voluntary and postural tasks.Participants performed a reaction time task. A Warning signal indicated to prepare to flex shoulders (postural, n=15) or to tilt the pelvis (voluntary, n=13) at the Go signal. Single- and paired-pulse TMS (short-interval intracortical inhibition - SICI) were applied using PA- and AP-TMS before the Warning signal (baseline), between the Warning and Go signals (preparation) or 30 ms before the LES onset (execution). Changes from baseline during preparation and execution were calculated in AP/PA-TMS.In the postural task, MEP amplitude was higher during the execution than preparation independently of the current direction (p=0.0002). In the voluntary task, AP-MEP amplitude was higher during execution than preparation (p=0.016). More PA-inhibition (SICI) was observed in execution than in preparation (p=0.028).Different neural circuits are preferentially involved in the two motor tasks assessed, as suggested by different patterns of change in execution of the voluntary task (AP-TMS: increase; PA-TMS: no change). Considering that PA-TMS preferentiallydepolarize neurons in M1,it questions their importance in LES voluntary control.Significance StatementBack muscles fulfill different roles as postural and control involving different neuronal circuits. Manipulating the electrical current direction induced by transcranial magnetic stimulation may allow the examination of different neural circuits contributions to postural and voluntary control of back muscles. In the execution of a postural task, corticospinal excitability was higher for both current directions than preparation. In the voluntary task, the corticospinal excitability was higher during execution than preparation using anteroposterior current only. Neural circuits contribution to back muscles control may depend on their role in the task performed. Our results suggest a minimal involvement of motor cortex neurons (minimally those interacting with posteroanterior current) in voluntary control of back muscles.