How Fast Could a Gallimimus Actually Run?

Most published estimates for Gallimimus’s top speed fall somewhere between 30 and 56 km/h, depending on the method used and the assumptions baked in. That range is wide enough to span the difference between a jogging human and a sprinting greyhound, which tells you something about the state of the science. Gallimimus was almost certainly one of the faster large dinosaurs, built with long, slender legs and a lightweight frame, but pinning down an exact number requires dealing with the fundamental problem that no one has ever watched one run.

An Animal Built for Speed

Gallimimus bullatus was an ornithomimid, a group whose name literally translates to “bird mimic.” Discovered in the Late Cretaceous deposits of Mongolia and described in 1972, it was one of the largest members of its family, stretching roughly six meters from snout to tail and weighing an estimated 400 to 500 kg. Picture something roughly ostrich-shaped but scaled up considerably, with a long neck, a toothless beak, and powerful hindlimbs that made up a large proportion of its total body length.

Several anatomical features point toward running ability. Gallimimus had an arctometatarsalian foot, a specialized arrangement where the middle metatarsal bone is pinched between the two flanking ones. This structure, shared with tyrannosaurs and other fast-moving theropods, is thought to have transmitted ground-reaction forces more efficiently during locomotion. Its lower leg bones were long relative to its femur, a ratio that in living animals generally correlates with cursorial habits. And its overall build was lightly constructed for a dinosaur its size, with thin-walled bones and a body plan that minimized mass in the extremities.

But anatomy alone does not give you a speedometer reading. It tells you the animal was built for some kind of sustained or rapid locomotion. Converting those skeletal proportions into an actual speed figure requires biomechanical modeling, comparative analysis, or fossil trackway data, each of which comes with its own set of problems.

Two Ways to Estimate a Dead Animal’s Speed

Researchers have two main toolkits for this question. The first is fossil trackways: preserved footprints left in mud or sand that sometimes capture stride length, foot shape, and gait. The second is biomechanical modeling, where researchers build digital or physical models of the animal’s skeleton and musculature and simulate how it might have moved. The two approaches often produce different numbers, and neither is without serious caveats.

The trackway approach dates to the mid-1970s, when R. McNeill Alexander derived a formula relating stride length, hip height, and speed using observations of living animals. He then applied this relationship to dinosaur trackways. The method is elegant in its simplicity, but it has a ceiling: Alexander himself warned that it “cannot claim to be accurate,” noting that the relationship between stride and speed might differ between dinosaurs and modern mammals, and that the leg-length estimates fed into the equation were themselves uncertain.¹ His original application to known dinosaur trackways produced estimated speeds between 1.0 and 3.6 meters per second, roughly the range from a slow walk to a brisk jog.² That tells us what those particular animals were doing at the moment their feet hit the mud, which was apparently nothing dramatic. It does not tell us the fastest they could have gone.

This is a crucial distinction. Trackways record what an animal happened to be doing, not what it was capable of. Most animals spend most of their time walking. Finding a preserved sprint is extremely rare, and even when faster trackways do turn up, interpreting them requires careful analysis of how foot posture changes at different speeds.

What Computer Simulations Suggest

Biomechanical modeling attacks the problem from the other direction. Instead of measuring what an animal did, it tries to calculate what it could have done, given the size and arrangement of its bones, the likely volume of its muscles, and the physics of ground contact. One influential approach used evolutionary robotics to build simple musculoskeletal models of bipedal animals, tested them against three living species to confirm the models could predict known top speeds, and then applied the same technique to five extinct species.³ The method produced estimates that matched real-world data for the living species reasonably well, which gives some confidence in the extinct-animal predictions, though “reasonable agreement” is doing heavy lifting in a field where you cannot ground-truth your results.

A separate modeling effort analyzed nine theropod dinosaurs spanning over five orders of magnitude in body mass. Its central finding was that running ability evolved in an inverse relationship to body size: smaller theropods were adept runners, while large theropods were not built for speed.⁴ Gallimimus sits in an interesting middle zone here. At 400 to 500 kg, it is well below the multi-ton mass of a tyrannosaur but substantially heavier than small theropods where speed estimates climb highest. The modeling results consistently suggest that body mass is the dominant variable controlling top speed, with anatomy playing an important but secondary role.

Why Limb Proportions Are Not a Simple Speedometer

For decades, paleontologists have used hindlimb proportions as a rough proxy for running ability. A longer shin relative to the thigh, a longer metatarsus relative to overall leg length: these ratios are associated with faster animals among living mammals and birds, and Gallimimus scores well on most of them. It is tempting to look at those proportions, compare them to an ostrich, and estimate a similar top speed.

The problem is that this shortcut falls apart once you account for body mass. A study that tested several common limb-proportion indices across a wide sample of theropods found that while distal limb proportions showed a modest correlation with estimated top speed on their own, correcting for body mass caused the correlation to collapse. The best-performing metric, the distal limb index, dropped to a correlation coefficient below 0.04 once mass was factored in.⁵ In plain terms, long legs in a small animal predict speed reasonably well; long legs in a large animal do not, because the mass penalty overwhelms the mechanical advantage.

This finding has significant implications for Gallimimus specifically. Its impressive leg proportions have been a major reason it has traditionally been placed at the faster end of dinosaur speed estimates. If those proportions are less predictive than previously assumed at its body mass, the higher speed estimates may be optimistic. Gallimimus was almost certainly quick for its size, but whether it was genuinely ostrich-fast is a different question.

The Problem of Posture

Another layer of uncertainty comes from how we reconstruct the way these animals actually held their legs while running. Modern birds run with a strongly flexed knee and a nearly horizontal femur, which keeps their center of mass over their feet. For years, many reconstructions of non-avian theropods used a similar posture. But the skeletal evidence suggests this is wrong for most dinosaurs.

Studies comparing limb-bone scaling between birds and non-avian theropods have found that the two groups are built quite differently. Non-avian theropods, including ornithomimids, have relatively long, slender femora that do not appear adapted to withstand the bending forces associated with a bird-like crouched posture.⁶ This implies that Gallimimus ran with a more upright leg than an ostrich does, which would have reduced the muscular effort needed to support its weight but also changed its stride mechanics in ways that affect speed calculations. A more upright limb is energetically cheaper for walking and moderate-speed running but does not generate as much forward thrust during an all-out sprint.

Biomechanical modeling supports this: the results strengthen the idea that many non-avian theropods, especially larger ones, used fairly upright limb orientations, which reduced the muscle mass required for locomotion.⁴ For Gallimimus, this means reconstructions that show it running in a deep-knee crouch are probably inaccurate, and speed estimates derived from those postures may need revision.

How the Tail Changed Everything

One of the more surprising recent findings in dinosaur locomotion research is how much the tail mattered. Living birds have greatly reduced tails, so their locomotion is driven almost entirely by the legs. Non-avian theropods, including Gallimimus, carried a long, muscular tail that shifted their center of mass further back relative to the hips. This changes the biomechanics of running substantially.

Experiments with chickens fitted with artificial tails that shifted their center of mass backward showed dramatic differences in how the animals moved. With the prosthetic tail in place, the femur’s range of motion during each step nearly tripled compared to normal chickens, going from about 15 degrees to almost 44 degrees.⁷ The tail-weighted chickens moved with a more retracted femur and greater hip-driven propulsion, which is probably a better analogue for how Gallimimus actually moved than a normal bird’s knee-driven stride.

This matters for speed estimation because hip-driven locomotion and knee-driven locomotion produce different stride characteristics, different ground-reaction forces, and different top speeds for the same leg length. Any model that assumes Gallimimus ran like a scaled-up ostrich without accounting for the tail’s influence on posture and kinematics is missing a major piece of the puzzle.

Running Fast Enough to Need a Warm Engine

One question that intersects with the speed debate is whether Gallimimus had the metabolic capacity to sustain fast running at all. Sprinting is enormously expensive energetically, and the ability to run at high speed for more than a few seconds requires a metabolic system that can deliver oxygen to muscles quickly.

Research estimating the locomotor metabolic rates of various bipedal dinosaurs has found that even moderate running speeds demanded metabolic output exceeding what any living cold-blooded animal can produce. At a moderate run, every species tested exceeded the aerobic ceiling for ectotherms, and at a slow run, all but the smallest species in the sample still exceeded it.⁸ This is strong indirect evidence that Gallimimus and its relatives were warm-blooded, or at least had metabolic rates much closer to birds and mammals than to modern reptiles. Without that elevated metabolism, the speeds we are discussing would have been physiologically impossible.

There is a counterpoint worth noting. A study of cardiopulmonary anatomy in theropods concluded that these animals were unlikely to have possessed a fully bird-like air-sac respiratory system, suggesting their lung efficiency may have been more comparable to crocodilians than to modern birds.⁹ If Gallimimus lacked the supercharged respiratory system of a modern bird, it may have been limited in how long it could sustain top speed, even if its leg mechanics were capable of impressive bursts. This could mean its running style was more like a short explosive sprint than the sustained high-speed travel ostriches are known for.

Clues from How Fast Theropods Actually Ran

While no Gallimimus trackway has been identified with certainty, some exceptionally fast theropod trackways from Spain offer a window into what running theropods actually did with their feet. Analysis of two of the fastest dinosaur trackways in the fossil record revealed something unexpected: the footprints varied between fully digitigrade (walking on the toes) and showing metatarsophalangeal impressions (the joint between the toes and the metatarsus touching the ground). The researchers interpreted these differences as reflecting dynamic changes in posture, weight distribution, and muscular activity during high-speed running.⁴

This is a genuinely new insight. It suggests that theropod running was not a simple, uniform gait but involved complex adjustments in foot posture, much the way a human sprinter’s foot contact changes between acceleration and top-speed phases. For Gallimimus, with its specialized arctometatarsalian foot, these postural shifts could have been especially pronounced. The pinched middle metatarsal may have played a role not just in overall limb stiffness but in how the foot transitioned between different phases of a stride at high speed.

What a Broken Bone Tells Us

One unexpected source of information about ornithomimid locomotion comes from pathology. A large-bodied ornithomimid, estimated at about 432 kg and thus roughly Gallimimus-sized, was found with a catastrophically fractured second metatarsal. The injury showed multiple oblique fractures consistent with blunt force trauma to the foot, followed by a bone infection. The fracture probably impaired the animal’s ability to bear weight on that foot until it died.¹⁰

The researchers suggested the injury could have resulted from interspecific or intraspecific competition, or from a predator-prey encounter. Whatever the cause, the pattern of damage is consistent with the kind of high-energy impact that would be expected in an animal moving fast or making sudden, forceful movements. A metatarsal fracture in a cursorial animal is roughly analogous to a racehorse breaking a cannon bone: it tells you the limb was under substantial mechanical stress, the kind of stress generated by high-speed locomotion or sudden maneuvers.

The fact that the animal survived long enough for the fracture to partially heal before infection set in also suggests something about ornithomimid biology. A large cursorial animal with a broken foot bone is extremely vulnerable to predation. Surviving for any period implies either that the animal could still move on three functional metatarsals, or that it had some social context that afforded protection. Either way, the foot was clearly a load-bearing structure subject to serious mechanical demands.

Where the Estimates Actually Land

Given all these methods and caveats, what can we actually say? The most honest framing is that Gallimimus could probably reach a top speed somewhere in the range of 30 to 50 km/h, with the lower end of that range increasingly favored by researchers who account for body-mass constraints and posture corrections. The higher estimates in older literature, some of which placed ornithomimids above 60 km/h, tended to rely on simple limb-proportion comparisons to ostriches without adequately correcting for the mass difference or the postural differences between birds and non-avian theropods.

For context, 35 km/h is roughly the speed of a fast recreational cyclist, and 50 km/h is about how fast an elite human sprinter moves during the fastest portion of a 100-meter dash. Gallimimus was almost certainly faster than any human who has ever lived and could likely have outrun most large predators in its ecosystem. Whether it could have matched an ostrich, which tops out around 70 km/h, is doubtful given its greater mass and less efficient respiratory system.

The ongoing refinement of biomechanical models, the discovery of new trackways, and advances in understanding theropod soft-tissue anatomy all continue to narrow the range. But the fundamental challenge remains: you cannot clock a dead animal. Every estimate is a model, and every model is a simplification of a living system whose full complexity is lost to time. The range of plausible speeds has tightened over the past few decades, which is progress. Expecting a single, confident number may be asking more than the fossil record can deliver.

The Jurassic Park Problem

It is impossible to discuss Gallimimus speed without acknowledging the scene that cemented the animal in popular culture: the stampede sequence in the original 1993 Jurassic Park film, where a herd of Gallimimus runs alongside a moving vehicle at what appears to be highway speed. The visual impression is of animals moving at 60 km/h or more, effortlessly matching a Jeep. This has anchored public perception of Gallimimus as one of the fastest dinosaurs, an impression that is not entirely wrong but is probably exaggerated.

Film animators were not constrained by biomechanical modeling. The Gallimimus in the movie run with a pronounced, ostrich-like gait that looks spectacular but may not reflect how the real animal moved. The speed, the stride, and the posture are all cinematic choices rather than paleontological reconstructions. The real animal was fast for a dinosaur. It was not, in all likelihood, as fast as a car.