The motor cortex comprises interconnected fields on the posterior frontal lobeâÂÂchiefly Brodmann area 4 (primary motor cortex, M1) and area 6 (premotor cortex and supplementary motor areas)âÂÂthat plan, select and execute voluntary movements. These regions transform goals into patterned activity in descending pathways to brainstem and spinal motor circuits, enabling dexterous eye, face and limb actions. Modern work shows overlapping, actionâÂÂtype representations rather than a strictly pointâÂÂtoâÂÂpoint "homunculus," and highlights direct corticoâÂÂmotoneuronal projections that underwrite fine finger control. Clinically, motorâÂÂcortical organization shapes deficits after stroke and neurodegenerative disease and guides mapping for neurosurgery and neurotechnology.
Motor cortex is commonly divided into three closely interacting fields:
In classical cytoarchitectonics, Brodmann area 4 (BA4) corresponds to primary motor cortex (M1) occupying the precentral gyrus and the anterior bank of the central sulcus, with medial continuation in the anterior (motor) portion of the paracentral lobule. Its posterior border abuts primary somatosensory cortex (BA3,1,2) along the lip and wall of the central sulcus; its anterior border is the precentral sulcus where area 6 begins. Receptorarchitectonic work subdivides BA4 into a posterior field (4p) concentrated along the sulcal wall and an anterior field (4a) on the gyral crown. Area 6 lies anterior to BA4 across the superior and middle frontal gyri and includes the lateral premotor cortex; on the medial wall it encompasses the supplementary and preâÂÂsupplementary motor areas.
Human and nonâÂÂhuman primate atlases differ in labeling schemes across anterior agranular cortex. In macaques, premotor fields are often subdivided into F2/F4 (dorsal/ventral caudal) and F5/F7 (ventral/dorsal rostral), which only partly correspond to human PMd/PMv. In humans, receptorarchitectonic divisions of BA4 into 4a/4p and probabilistic maps derived from imaging produce slightly different borders than gyral/sulcal landmarks, especially near the central sulcus.
M1 contains large pyramidal neurons (Betz cells) in layer V and projects densely to spinal and cranial motor circuits via the corticospinal and corticonuclear tracts. Although Betz cells are distinctive, they form only a small proportion of corticospinal outputs; most corticospinal fibers arise from nonâÂÂBetz layer V neurons in M1 and from adjacent motor areas.
Premotor cortex is commonly divided into dorsal (PMd) and ventral (PMv) sectors, each with rostral and caudal parts. PMd contributes to reach planning and selection among competing directions, whereas PMv is heavily involved in shaping the hand for grasp and in multisensory guidance of actions in periâÂÂpersonal space. These areas are part of a broader parietoâÂÂfrontal system linking dorsal visual streams with motor plans, and their boundaries lie within cytoarchitectonic area 6 lateral to BA4.
In macaque PMv (area F5), some neurons fire both during execution of a grasp and during observation of the same action performed by others; these "mirror" responses have been proposed to contribute to action understanding and imitation. The extent and function of mirrorâÂÂlike responses in humans remain debated, but convergent EEG/MEG and fMRI evidence shows actionâÂÂobservation effects in premotor and parietal circuits that project to M1.
The frontal eye field (FEF) in the precentral/premotor region and the supplementary eye field (SEF) on the dorsomedial wall form part of the motor network controlling saccades, smooth pursuit and eyeâÂÂhead coordination. FEF receives visual input from occipitoâÂÂtemporal pathways and projects to the superior colliculus and brainstem gaze centers; SEF participates in internally generated saccade sequences and performance monitoring. Microstimulation of FEF evokes fixedâÂÂvector saccades, whereas SEF stimulation elicits contextâÂÂdependent eye movements and sequence effects.
Electrical stimulation and functional imaging implicate SMA in initiating internally generated action and in sequencing. SMA also contains a coarse, overlapping body map and sends direct corticospinal projections. Lesions or inactivation can impair movement initiation and transiently abolish bimanual coordination in nonâÂÂhuman primates.
Motor cortex is agranular isocortex with a sixâÂÂlayered structure; layer IV is reduced or indistinct, whereas layer V contains the large corticospinal neurons. M1 is sometimes termed area gigantopyramidalis because Betz cells are especially prominent there. Premotor and SMA share a similar laminar pattern but lack Betz cells. Afferent input arrives via thalamic relays conveying basal ganglia and cerebellar output; rich corticocortical connections link PMd/PMv with posterior parietal cortex and SMA with prefrontal cortex. Efferents descend via the corticospinal and corticonuclear tracts and via brainstem motor pathways.
M1, premotor cortex and SMA are agranular isocortex. Layer IV is attenuated or indistinct, while layer V contains large pyramidal neurons including Betz cells in M1. Neuron classes include corticospinal and corticobulbar projection neurons, corticocortical pyramidal cells in layers II/III and V, and diverse GABAergic interneurons (basket, chandelier, Martinotti). Cortical thickness varies across the precentral gyrus from the gyral crown to the anterior sulcal wall, paralleling shifts in input/output density and myelination. Betz cells constitute a small minority of corticospinal neurons but have exceptionally thick axons and fast conduction velocities.
Motor cortical output travels in the corticospinal tract (pyramidal tract) and corticobulbar systems. Fibers originate from multiple fields: approximately one quarter from small pyramidal neurons in M1, substantial fractions from premotor and SMA, and a sizable minority from somatosensory cortex; Betz cells account for only a few percent of corticospinal axons. Many corticospinal terminals contact spinal interneurons, whereas direct corticoâÂÂmotoneuronal connections are thought to underlie fine finger control.
Corticobulbar projections from lateral M1 and ventrolateral premotor cortex target cranial motor nuclei through relay zones in the pontine and medullary reticular formation. Orofacial, laryngeal and tongue representations occupy the inferior precentral gyrus and adjacent opercular cortex. Direct corticoâÂÂmotoneuronal influences on nucleus ambiguus (laryngeal) are sparse in most mammals but appear more substantial in humans and great apes, consistent with fine control of phonation and articulation. Lesions produce dysarthria and apraxia of speech; stimulation studies and functional imaging localize laryngeal motor cortex to a dorsalâÂÂventral pair flanking the central sulcus.
Rather than oneâÂÂtoâÂÂone control of individual muscles, stimulation and singleâÂÂunit studies indicate that motor cortex contains heavily overlapping representations and can specify ethologically relevant, multiâÂÂjoint actions. ExtendedâÂÂduration microstimulation in monkeys evokes coordinated movements such as defensive postures or reachâÂÂtoâÂÂgrasp sequences, suggesting a map of action types arranged across cortex.
Population vectors, directional tuning and dynamicalâÂÂsystems descriptions have been used to account for how ensembles in motor cortex evolve during reach and grasp. BetaâÂÂband (âÂÂ13âÂÂ30 Hz) oscillations increase during hold periods and desynchronize around movement onset; highâÂÂgamma activity scales with force and kinematics in electrocorticography.
Scalp and intracranial recordings show a slow negative potential, the readiness potential, beginning up to 1âÂÂ2 s before selfâÂÂinitiated movement. Sources include SMA, preâÂÂSMA and M1, with lateralized readiness potentials reflecting effector selection.
Several accounts describe how motor cortex specifies movement: (i) muscleâÂÂbased coding, in which neurons correlate with muscle activity; (ii) movementâÂÂbased coding, emphasizing kinematics/forces of effectors; and (iii) dynamicalâÂÂsystems views, in which population activity flows along lowâÂÂdimensional trajectories that generate movement without requiring an explicit setâÂÂpoint for each muscle. Optimal feedback control frames motor behavior as taskâÂÂlevel goals stabilized by feedback, with motor cortex participating in a distributed controller.
Motor representations are shaped by development and use. Early corticospinal projections are exuberant; activityâÂÂdependent pruning and myelination refine conduction velocity and terminal specificity through childhood and adolescence. Experience can expand or contract cortical zones devoted to particular movements, and recovery after injury may recruit premotor and somatosensory contributions to descending pathways.
Transient bilateral projections are common in infancy; progressive myelination, synaptic pruning and strengthening of corticoâÂÂmotoneuronal connections accompany the emergence of fine manual dexterity. Diffusion MRI and TMS demonstrate increasing tract integrity and decreasing motor thresholds across childhood and adolescence.
Skill acquisition alters representational geometry in M1 and premotor cortex, biases corticoâÂÂmotoneuronal drive toward task muscles, and modifies intracortical inhibition/facilitation. NonâÂÂinvasive stimulation (e.g., TMS, tDCS) can transiently modulate learning rates and retention.
Aging is accompanied by altered recruitment of premotor and contralateral homologues during motor tasks and by changes in myelination and thickness gradients across precentral cortex; training can partially normalize these patterns.
Across mammals, corticospinal organization varies with dexterity. Species with skilled, independent finger movements (e.g., humans, macaques) possess abundant corticoâÂÂmotoneuronal projections and a prominent M1 âÂÂhand knob,â whereas species with less manual dexterity rely more on propriospinal and brainstem pathways. In nonâÂÂprimates, corticospinal fibers terminate largely on interneurons, while in higher primates many terminations contact motoneurons directly. The distribution and strength of CM projections correlate with the capacity for independent finger movements and tool manipulation.
Tool use and fine object manipulation in primates rely on parietoâÂÂfrontal networks linking anterior intraparietal areas with ventral premotor cortex and M1. Expansion of these circuits in humans is associated with increased CM projection density and greater fractional representation of distal musculature, supporting skilled grasp, tool use and praxis.
Preoperative and intraoperative mapping reduce morbidity in resections near motor areas. Direct cortical stimulation during awake craniotomy identifies essential sites for movement and speech articulation; subcortical stimulation traces the course of descending fibers. Outside the operating room, transcranial magnetic stimulation (TMS) defines resting motor threshold, motor maps, and intracortical inhibition/facilitation using pairedâÂÂpulse protocols.
Intraoperative mapping commonly uses short trains of biphasic or monophasic pulses delivered via bipolar electrodes placed on the cortical surface (typical frequencies ~50âÂÂ60 Hz; train durations on the order of 1âÂÂ5 s; currents in the low milliampere range adjusted to evoke responses while avoiding afterdischarges). Mapping proceeds in small spatial steps to delineate essential cortex and whiteâÂÂmatter pathways; electrocorticography is used to monitor afterdischarges, and stimulation is paused or medication given if they arise.
Resting motor threshold (RMT) is defined as the minimum stimulator output evoking motor evoked potentials (MEPs) of standard amplitude in a proportion of trials at rest; active motor threshold (AMT) is defined during slight contraction. PairedâÂÂpulse TMS quantifies shortâÂÂinterval intracortical inhibition and intracortical facilitation, while mapping protocols estimate the areal extent of corticomotor representations.
Focal lesions of M1 produce contralateral weakness, loss of fine fractionated movements and pathological reflexes, whereas lesions affecting premotor and SMA tend to impair movement initiation, selection, praxis and bimanual coordination. Symptoms reflect the somatotopic bias of affected territories and the crossing of corticospinal fibers in the medullary pyramids.
Upper motor neuron signs (spasticity, hyperreflexia, Babinski sign) reflect corticospinal and cortical dysfunction. In amyotrophic lateral sclerosis (ALS), degeneration involves layer V corticospinal neurons as well as spinal motor neurons; neurophysiology often shows corticospinal hyperexcitability early in disease. Imaging and pathology implicate precentral gyrus atrophy and corticospinal tract degeneration.
Primary lateral sclerosis presents with progressive upper motor neuron involvement and spastic paraparesis or quadriparesis with relative sparing of lower motor neurons; abnormalities of motor cortex excitability and corticospinal conduction are typical. Corticobasal syndrome features asymmetric apraxia, rigidity and dystonia linked to frontoparietal degeneration; premotor and parietal dysfunction contribute to impaired praxis and alienâÂÂlimb phenomena.
Ischemic injury affecting the precentral gyrus or its descending fibers causes hemiparesis or hemiplegia. Recovery engages perilesional M1, ipsilesional premotor cortex, contralesional homotopic areas and cerebelloâÂÂthalamoâÂÂcortical circuits. TaskâÂÂoriented therapy and neuromodulation aim to bias adaptive plasticity.
Abnormal excitability and loss of intracortical inhibition within motor cortex contribute to spasticity, dystonia and taskâÂÂspecific cramps. Focal cortical dysplasias in the precentral region may cause motor seizures.
Lesions of lateral premotor areas and inferior parietal cortex can produce ideomotor apraxia. Medial wall resections or stroke involving SMA often yield transient akinesia and impaired bimanual coordination (SMA syndrome) with gradual recovery.
Electrical stimulation studies by Fritsch and Eduard Hitzig (1870) and by David Ferrier (1870s) first demonstrated a motor representation in cortex. Experiments by Hermann Munk and contemporaries debated sensory vs. motor localization. In the 20th century, intraâÂÂoperative stimulation in humans by Penfield and colleagues produced the iconicâÂÂthough simplifiedâÂÂâÂÂhomunculus,â while later work emphasized distributed and overlapping motor maps. Contemporary studies integrate microstimulation, singleâÂÂunit and population recordings, and neuroimaging to refine these maps.
Recordings from motor cortex have been used to control external devices in proofâÂÂofâÂÂconcept brainâÂÂcomputer interface (BCI) studies in humans with tetraplegia, demonstrating realâÂÂtime control of cursors and robotic limbs.
Motor cortex has been investigated with lesions, electrical stimulation, intracortical microstimulation, singleâÂÂunit and local field recordings, electrocorticography, MEG/EEG, fMRI, TMS/tDCS and pharmacologic perturbations. Each technique samples different spatial and temporal scales and has distinct interpretive limits.
Spatial precision, sampling, and signalâÂÂtoâÂÂnoise differ markedly across modalities (e.g., fMRIâÂÂs indirect hemodynamic signal vs. ECoGâÂÂs highâÂÂtemporalâÂÂresolution field potentials). Mapping results depend on behavioral context and analysis assumptions; causal inference generally requires stimulation or lesion methods. NonâÂÂinvasive stimulation has safety and interpretive constraints summarized in consensus guidance.
Early electrical stimulation studies established a cortical motor center: in 1870, Fritsch and Hitzig showed that stimulating parts of dog cortex produced contralateral movements, and FerrierâÂÂs subsequent work in primates mapped a rough dorsoventral progression from leg to face. Clinical stimulation and lesion studies in humans, popularized by Penfield midâÂÂ20th century, refined the human somatotopic picture and its limits.
Debate over field boundaries (area 4 vs. area 6) persisted into the 20th century; by the 1930sâÂÂ1950s a consensus emerged that cytoarchitectonic area 4 (M1) and area 6 (premotor/SMA) are functionally distinct but connected fields with partly overlapping descending outputs.
Contemporary work divides premotor cortex into dorsal and ventral sectors, each with rostral and caudal parts (PMDr, PMDc, PMVr, PMVc), largely based on architectonic, connectivity, and stimulation criteria in macaques. PMDc neurons are strongly engaged during reach preparation and execution, and microstimulation there can evoke coordinated shoulderâÂÂarmâÂÂhand movements resembling reaching. PMVr (classically F5) contains graspâÂÂrelated neurons and was the original site where âÂÂmirrorâ discharge was described in macaques; electrical stimulation on behavioral time scales in precentral regions can elicit complex, ethologically meaningful actions (e.g., defensive and handâÂÂtoâÂÂmouth sequences), underscoring that motor fields encode coordinated actions rather than single muscles.