Can Humans Run on All Fours? The Science Explained

Humans can move on all fours, and most of us did it as infants, but the adult body is so thoroughly redesigned for upright walking that quadrupedal locomotion becomes remarkably costly. Walking on hands and feet burns roughly two and a half times more energy than bipedal walking at the same speed, and the fastest human quadrupedal sprinters are still far slower than bipedal runners. The question of whether we can do it is less interesting than why our bodies fight it so hard, and what happens when people try anyway.

An Anatomy Built Against It

The human skeleton is not simply a quadruped’s skeleton stood upright. Millions of years of selection for bipedalism reshaped almost every load-bearing structure from the skull down. The pelvis is the most obvious example: in our earliest upright ancestors, the pelvis became shorter and wider compared with other primates, repositioning the gluteal muscles so they could stabilize the trunk over a single supporting leg during each stride.¹ That pelvic shape, combined with the lumbar curve that develops when an infant learns to walk, creates a tightly integrated pelvis-and-spine unit tuned specifically for an upright gait cycle.²

Limb proportions tell the same story. Among primates, the ratio of arm length to leg length, called the intermembral index, is the single best anatomical variable for distinguishing great apes and monkeys from one another and from humans.³ Humans have unusually long legs relative to their arms. When you drop onto all fours, your hips sit much higher than your shoulders, tilting your trunk steeply downward. That angle forces your wrists to bear loads they are not shaped for and puts your spine in a position that fights its natural curvature.

The shoulder is another problem. In quadrupedal primates, the shoulder blade is long and oriented in line with the spine so that it can channel the forces of each forelimb footfall straight through the trunk. Human shoulder blades, by contrast, sit more toward the side of the rib cage with a shape adapted for overhead reaching and suspension rather than ground-level weight-bearing.⁴ Trying to walk on your hands loads the shoulder in a direction it was not built to handle repeatedly.

The Energy Penalty

If anatomy alone does not convince you, the metabolic numbers will. A 2024 study measured the energy cost of quadrupedal versus bipedal walking in healthy adults moving at the same speed. Quadrupedal walking increased total energy expenditure by about 4.15 kilocalories per minute over bipedal walking, a jump driven largely by a spike in carbohydrate burning. Heart rate and breathing rate both climbed sharply. Over just ten minutes at the same pace, quadrupedal walking consumed roughly 254% more energy than walking upright.⁵

That gap makes intuitive sense once you consider how the body has to compensate. Your arms are doing propulsive and weight-bearing work they normally avoid. Your core muscles fire in unfamiliar patterns. And because your legs are longer than your arms, each stride involves an awkward compromise: either you bend your knees deeply to keep your hips low, wasting energy on maintaining a crouch, or you keep your legs straighter and let your trunk angle steeply, loading your wrists and shoulders. Neither option is efficient. The researchers behind the energy study noted that their results support the long-standing idea that energy optimization was a major evolutionary driver of bipedalism: walking upright is simply a better deal, calorie for calorie.⁵

How the Spine Handles Quadrupedal Loads

Bipedal walking and quadrupedal walking demand very different things from the muscles that stabilize the spine. Comparisons between human and sheep trunk-muscle activity reveal that during walking, humans use significantly more abdominal muscle activity than quadrupeds do, while quadrupeds rely more heavily on their spinal extensor muscles.⁶ During running and trotting, the pattern diverges even further: humans show a greater range of trunk-muscle activation across nearly all muscle groups.

What this means for someone trying to move on all fours is that the human trunk-stabilization system is wired for an upright posture. When you flip to a horizontal position, the muscles that normally keep your torso from collapsing forward, especially the deep abdominal layers, are suddenly asked to do a job they have been trained for only in an upright context. The extensor muscles along the spine, which in a habitual quadruped are the primary stabilizers, are comparatively weaker in that role in humans. The result is a spine that wobbles, fatigues quickly, and complains.

How Fast Can a Human Sprint on All Fours?

There is an actual competitive scene for this. Quadrupedal sprinting has been recorded in organized events, and researchers have tracked the progression of 100-meter times over the years. A study published in Frontiers in Bioengineering and Biotechnology analyzed both bipedal and quadrupedal world-record times and fitted trend lines to them. They found that quadrupedal sprint records have been improving at a faster rate than bipedal ones, and the two curves are projected to intersect around the year 2048, when a hypothetical quadrupedal 100-meter time of about 9.28 seconds would edge out a projected bipedal time of about 9.38 seconds.⁷

Before you take that projection too seriously, some context. The researchers were explicit that this is a mathematical extrapolation, not a prediction. The improvement curve for quadrupedal sprinting is steep partly because the sport is young and attracting new athletes who push times down rapidly, a pattern that typically flattens as any sport matures. The current quadrupedal 100-meter record is nowhere near 9.28 seconds. Still, the study is useful because the video analysis it included reveals something about technique: human quadrupedal runners use a transverse gallop, a gait where the two forelimbs and two hindlimbs land in a lateral sequence. The researchers noted that the angular excursion of the trunk during this gallop is small, and suggested that faster times might come from shifting to a rotary gallop with wider trunk motion and longer strides.⁷

In other words, humans on all fours are still learning how to gallop. Quadrupedal animals have had millions of years to refine their gaits, while human quadrupedal runners are essentially inventing a movement pattern from scratch. The fact that they are improving quickly says more about the novelty of the activity than about any hidden quadrupedal potential in human anatomy.

What Your Nervous System Does When You Crawl

One of the more fascinating aspects of human quadrupedal movement is what it reveals about the nervous system. Your spinal cord contains networks of neurons called central pattern generators that can produce rhythmic limb movements without constant input from the brain. These generators are separate for the arms and legs, but they communicate with each other through connections running up and down the spinal cord.

When healthy adults crawl on all fours, the coordination between upper and lower limbs tends to lock into specific frequency ratios, usually moving the arms and legs at the same rate or at simple integer multiples of each other. This coupling pattern is evidence that the spinal-cord generators for the arms and legs are linked. But the link is not symmetrical. The arms seem more willing to break away from a locked rhythm than the legs, especially at slower speeds. Researchers have speculated that this asymmetry exists because the connections running upward from the leg generators to the arm generators are weaker than the descending ones, and because the brain’s cortex exerts more direct control over the arms, allowing it to override the spinal rhythm more easily.⁸

This makes sense from an evolutionary standpoint. Humans freed their hands from locomotion, and the nervous system adapted by giving the cortex tighter control over the arms. When you crawl, you are asking a system optimized for independent hand use to fall back into a rhythmic, locomotor pattern. It can do it, but the coordination feels less automatic than walking on two legs.

Uner Tan Syndrome and Habitual Quadrupedalism

The most striking real-world examples of humans moving on all fours come from a rare condition called Uner Tan syndrome. First described in several families in rural Turkey, the syndrome involves habitual quadrupedal locomotion on palms and feet, along with cognitive impairment and speech difficulties. Brain scans of affected individuals typically show underdevelopment of the cerebellum, the brain region critical for balance and motor coordination.⁹

Genetic investigation has identified mutations in the VLDLR gene, which codes for a receptor involved in the signaling pathway that guides developing neurons to their correct positions in the brain. In two of the affected Turkish families, different mutations in VLDLR were found to produce truncated, nonfunctional versions of the protein, disrupting normal brain development and leaving the cerebellum undersized.¹⁰ The condition is genetically heterogeneous, meaning different mutations in different genes can produce the same outcome, and all known families have a history of consanguineous marriages, consistent with autosomal recessive inheritance.⁹

Uner Tan syndrome generated intense debate when it was first reported. Some commentators initially framed it as an “evolutionary throwback,” as if affected individuals had somehow reverted to an ancestral gait. That interpretation does not hold up. The quadrupedal gait seen in these families is not the same as the gait of any nonhuman primate; it is a compensatory solution that the developing nervous system arrives at when the cerebellum cannot support upright balance. Researchers who studied the syndrome concluded that the quadrupedalism likely emerges through self-organization, brain plasticity, and rewiring rather than through the reactivation of some dormant ancestral program.⁹ The body finds a way to move given the hardware it has. It is an impressive demonstration of neural adaptability, not a window into our quadrupedal past.

Quadrupedal Movement as a Fitness Tool

Outside the lab and the medical literature, quadrupedal movement has found a niche in fitness training. “Animal flow,” bear crawls, and other four-limbed ground exercises have become popular in functional fitness programs. The appeal is straightforward: moving on all fours loads the shoulders, wrists, hips, and core in ways that upright exercises do not, and it demands coordination across all four limbs simultaneously.

A controlled trial tested whether a structured quadrupedal movement training program could improve functional movement, range of motion, and strength. After the training period, the quadrupedal group showed significantly greater improvements in overall functional movement scores compared with a control group, along with meaningful gains in hip flexibility and shoulder range of motion.¹¹ The researchers concluded that quadrupedal movement training is a viable alternative for improving whole-body stabilization and flexibility.

This makes physiological sense. The energy penalty of quadrupedal walking, discussed earlier, is a feature rather than a bug when the goal is exercise. Burning roughly two and a half times more calories at the same speed means you get more metabolic stimulus in less time. And because the movement pattern is unfamiliar, it challenges motor coordination in ways that running or cycling on autopilot does not. The downside is that loading the wrists and shoulders with body weight can aggravate existing joint problems, so anyone with wrist tendinitis or shoulder impingement should approach it cautiously.

Head Stability and the Gaze Problem

One underappreciated challenge of quadrupedal locomotion in humans is what happens to your head. When you run upright, a neuromechanical linkage between the head and the forearm, mediated by coordinated activation of the biceps and the upper trapezius, helps stabilize the head and counteract the forward pitching that each footstrike produces.¹² This linkage is active during running but not during walking, which suggests it evolved specifically to keep your gaze steady at higher speeds.

On all fours, the situation is entirely different. Your head hangs downward and forward, and the muscles that stabilize it in upright running are now tasked with holding it up against gravity while also being involved in forelimb propulsion. The result is a conflict: the same muscles needed to look where you are going are also needed to move you forward. Quadrupedal animals solve this with different neck anatomy and differently positioned eyes. Humans have a face built for looking straight ahead while upright, and bending the neck backward far enough to see the horizon while on all fours is both uncomfortable and unsustainable over long periods. This is not just a matter of muscle fatigue. Your vestibular system, the inner-ear apparatus that tells you which way is up, is calibrated for a head that sits on top of a vertical spine. Tipping the whole system forward while trying to hold the head level introduces conflicting signals that can cause dizziness and disorientation, especially during faster movement.

Why Quadrupedal Gaits Look So Different in Humans

If you watch footage of quadrupedal sprinters, one thing immediately stands out: they do not move like dogs, horses, or cheetahs. The gait looks improvised, because it is. Quadrupedal animals fall into predictable gait categories, with footfall timing that follows well-defined patterns related to speed and duty factor, the fraction of each stride during which a given foot is on the ground. At slow speeds and high duty factors, most quadrupeds walk with diagonal limbs moving nearly together; at faster speeds and lower duty factors, they shift to trots and gallops with specific phase relationships between limbs.¹³ These patterns exist because they minimize the mechanical work of locomotion for a given body plan.

Humans trying to gallop cannot slot into these optimized patterns because the body plan is wrong. Arms and legs are different lengths. The hip joint is oriented for extension in the sagittal plane, not for the kind of flexion-extension cycling a quadruped’s hindlimb performs. The wrist is a precision instrument for manipulation, not a load-bearing strut. So human quadrupedal runners end up with a transverse gallop characterized by limited trunk motion, essentially a workaround that avoids the most painful joint positions at the cost of stride length and speed. The researchers who analyzed quadrupedal sprinting video explicitly noted that athletes could improve by adopting a rotary gallop with wider trunk motion, but achieving that may be limited by how far the human spine and pelvis can rotate in a horizontal position without injury.

All of this points to a clear picture. Humans can run on all fours in the same way that a car can technically drive on three wheels: the hardware permits it, but nothing about the design encourages it. Every structural, metabolic, and neurological system in the human body has been shaped by millions of years of bipedal specialization. When we go back to four limbs, we are working against our own engineering, which is exactly what makes it such a demanding workout and such a poor way to get anywhere quickly.