Ridge push and slab pull are the two main gravitational forces that drive tectonic plates across Earth’s surface. Ridge push is the outward force generated at mid-ocean ridges, where hot mantle material rises, creates new seafloor, and the elevated ridge literally pushes older lithosphere away. Slab pull is the downward tug exerted by a cold, dense oceanic plate as it sinks into the mantle at a subduction zone. Of the two, slab pull is considerably stronger, but how these forces interact with each other and with resistance from the surrounding mantle shapes nearly everything about how plates move, how fast they go, and why Earth’s surface looks the way it does.
How Ridge Push Works
A mid-ocean ridge sits higher than the surrounding seafloor because the mantle beneath it is hot and buoyant. As new oceanic crust forms at the ridge and moves away, it cools, thickens, and becomes denser. The result is a gently sloping surface: high and warm near the ridge crest, low and cold far from it. Ridge push is essentially the force of gravity acting on that slope. The lithosphere slides down its own sloping base, driven by the elevation difference between the ridge crest and the deep abyssal plains thousands of kilometers away.
The force depends on how thick the plate has become and how much higher the ridge crest sits above the surrounding ocean floor. A thicker, older plate that has had more time to cool produces a steeper slope, which translates to a stronger push. This is a continuous, steady force that acts across the entire trailing plate, not just right at the ridge itself. Think of it less like a conveyor belt shoving plates from behind and more like a plate sliding down a very gentle, very long ramp.
Ridge push is not uniform around the globe. Spreading rates vary enormously, from around 150 millimeters per year at the fast East Pacific Rise to nearly zero at ultraslow ridges like the Gakkel Ridge in the Arctic. Mantle temperature beneath ridges also varies, dropping by more than 150°C from fast-spreading to ultraslow-spreading settings. Hotter mantle produces more melt and a taller ridge, which means a stronger push. So a fast-spreading ridge in the Pacific generates a more substantial ridge push force than a sluggish one in the Arctic.
How Slab Pull Works
When an oceanic plate reaches a subduction zone, it bends downward and sinks into the mantle. Because the plate is cold and dense compared to the surrounding hot mantle, gravity pulls it down. That downward pull is transmitted back along the plate like tension in a rope, dragging the trailing surface plate toward the trench. This is slab pull, and it is the single largest force in the plate-tectonic system.
What makes a slab heavy enough to sink is more than just temperature. As oceanic crust descends to depths of roughly 100 to 200 kilometers, low-density minerals in the basaltic crust transform into a much denser rock called eclogite. This mineral transformation adds significant weight to the sinking slab and increases the pull force. But the densification of the slab’s mantle portion also matters and actually begins earlier in the subduction process than the crustal transformation does.
The length and angle of the slab both affect how strong the pull is. A longer slab hanging deeper into the mantle has more mass pulling downward. A steeper dip angle concentrates that weight more directly beneath the trench. The net slab pull force, after accounting for resistance from the mantle around it, can reach roughly 8 to 12 percent of the total gravitational body force on the slab, which still makes it up to twice as large as ridge push.
The strength of the slab itself also matters. A slab is not infinitely rigid; it bends and can weaken as it enters the trench. If the connection between the deep slab and the surface plate is weak, less of the pull force actually reaches the trailing plate. This is why how well a slab transmits force back to the surface plate is just as important as how much force the sinking portion generates.
Which Force Dominates
Early quantitative work on plate-driving forces established that slab pull is roughly an order of magnitude stronger than any other force acting on the plates. A landmark study by Forsyth and Uyeda found that every oceanic plate attached to a substantial amount of sinking slab moves at a “terminal velocity” where the gravitational pull on the slab is nearly balanced by resistance from the mantle. The other features of the plate, including how much ridge it has, barely mattered for determining speed.
More recent modeling confirms this hierarchy. The net slab pull force, once mantle resistance is subtracted, still ends up roughly twice as large as ridge push. That means slab pull is the primary engine, and ridge push acts more like a secondary assist. In regions where no subduction is happening, ridge push becomes the dominant available force, but it produces much slower plate motion on its own.
Observational evidence lines up with this. The fastest-moving plates on Earth, clocking about 8.5 centimeters per year, have little continental material and are bounded by subduction zones. The slowest plates, moving around 2.6 to 2.8 centimeters per year, carry large continental fractions and have little or no subducting perimeter. Oceanic plates generally move two to three times faster than continental plates, exactly what you would expect if the dominant force comes from sinking slabs rather than from ridges.
Plates Without Subduction Zones
If slab pull is the main driver, how do plates without any subducting edges move at all? The African plate, for instance, is almost entirely surrounded by mid-ocean ridges and is not being dragged into any major subduction zone. It still moves, just slowly.
Ridge push provides part of the answer. Studies of the Indo-Australian plate show that ridge push exerts a substantial torque on the plate, and the direction of that torque closely matches the plate’s actual direction of motion. The intraplate stress field within continental Australia, including north-south compression in the north and east-west compression in the south, can be explained largely by ridge push forces balanced against resistance at the plate’s collisional boundaries in the Himalayas, Papua New Guinea, and New Zealand. Similar analysis for North America, western Europe, and South America shows that ridge push forces align well with the orientations of maximum stress measured inside these continents.
So while ridge push cannot compete with slab pull for raw strength, it plays an outsized role in determining the stress patterns and slow motions of plates that lack a subducting edge. For plates like North America, however, modeling suggests that active drag from deeper mantle flow may also contribute. One study found that only models with active drag from the lower mantle, rather than models where the mantle passively resists or exerts no force, successfully matched the observed spreading rates, stress directions, and geodetic velocities of the North American plate.
Slab Suction and Mantle Drag
Ridge push and slab pull are not the only forces at play. As a slab sinks, it displaces mantle material around it, setting up flow patterns that pull surrounding plates toward the subduction zone. This effect, called slab suction, acts on both the subducting plate and the overriding plate symmetrically, drawing them together. Unlike slab pull, which tugs only on the plate physically attached to the slab, slab suction works through the flowing mantle and can influence plates that are not directly connected to any sinking slab.
The importance of slab suction has shifted over Earth’s history. During the Cenozoic, the balance between direct slab pull and slab suction changed as subduction geometries evolved. Slab suction appears to become more significant when slabs detach from their surface plates, because the detached slab continues sinking and driving mantle flow even though it is no longer physically pulling on anything at the surface.
Mantle drag, where flowing mantle beneath a plate either resists or assists its motion, adds another layer of complexity. Traditionally, mantle drag was thought to resist plate motion, acting like friction on the base of the plate. But in some settings it actively drives plates forward. The North American plate, for example, appears to be pushed westward in part by lower-mantle flow. The picture that emerges is a system where slab pull and ridge push set the main direction and speed, while mantle drag and slab suction fine-tune the motion depending on local conditions.
What Happens When a Slab Breaks Off
Slabs do not always remain intact. A slab can tear or break off entirely at depth, a process called slab detachment or slab break-off. When this happens, the slab pull force on the trailing plate drops suddenly. You might expect the plate to stop converging, or at least slow down dramatically. But observations show that plates can keep converging even after a slab break-off.
The Cyprian Arc in the eastern Mediterranean provides a case study. Research on this subduction zone found that convergence continued after a slab break-off event, suggesting that other forces compensated for the lost pull. One mechanism for this is that slab detachment changes the pattern of mantle flow beneath the plates, transforming the shear traction at the base of the plates from a resisting force into a driving one. In other words, the detached slab keeps sinking and stirring the mantle, and that mantle flow can partially replace the lost direct pull.
This resilience is part of why plate tectonics does not simply shut down when individual slabs fail. The system has multiple overlapping force sources, and the loss of one can be partially offset by changes in the others.
Stalling in the Mantle Transition Zone
Slab pull does not operate uniformly all the way to the bottom of the mantle. At depths of about 410 and 660 kilometers, minerals undergo phase transitions that change the slab’s density relative to its surroundings. The 660-kilometer boundary is particularly important because conditions there can resist the slab’s downward progress. Some slabs punch through into the lower mantle, while others flatten out and stagnate in what geophysicists call the mantle transition zone.
Whether a slab stagnates or penetrates depends on several factors: older, colder slabs subducting at shallow angles with significant trench retreat tend to stagnate, while other configurations allow penetration into the lower mantle. Seismic imaging reveals both behaviors in different subduction zones around the world. A stagnating slab still exerts some pull on the surface plate, but the geometry changes how that force is transmitted. A slab lying flat at 660 kilometers pulls differently than one plunging steeply to 1,000 kilometers.
This variation helps explain why not all subduction zones behave identically. The Tonga trench, where the Pacific Plate dives steeply into the mantle, produces different plate speeds and stress patterns than the Cascadia subduction zone, where the Juan de Fuca plate subducts at a shallower angle. The deep structure of the slab matters for what the surface plate feels.
How Subduction Gets Started
One of the trickier questions in plate tectonics is how subduction initiates in the first place. A cold, dense plate sitting at the surface should, in principle, resist bending downward because the lithosphere is stiff. Some external push or internal weakness has to overcome that initial resistance before slab pull can take over.
Modeling of the Western Pacific, where the oldest oceanic plates on Earth exist, suggests that internal vertical forces can do the job without much horizontal pushing. In one study, researchers found that with the net initial horizontal push set to zero, a 2D representation of the Pacific-Philippine Sea plate system was already close to subduction initiation. Only a small additional vertical pull of about 9 teranewtons per meter was needed, far less than the roughly 30 teranewtons per meter typical of a mature slab pull force, for subduction to rapidly begin and progress to full down-dip sinking. Old, dense oceanic lithosphere sitting next to younger, more buoyant lithosphere at a fracture zone or transform fault can be enough to tip the balance.
Once subduction starts and a slab begins forming, the system becomes self-reinforcing. As more plate sinks, slab pull grows, which pulls more plate into the trench, which makes the slab longer, which increases the pull. Getting the process started is the hard part; sustaining it is comparatively easy.
The Supercontinent Connection
Ridge push and slab pull also factor into the grand cycles of supercontinent assembly and breakup that have punctuated Earth’s history. After a supercontinent like Pangaea breaks apart, ridge push from newly formed mid-ocean ridges helps scatter the continental fragments outward, while slab pull at the leading edges of oceanic plates draws them toward subduction zones. These two forces work in concert during the dispersal phase.
Reassembly is more complicated. In one scenario called introversion, the scattered continents reverse course and come back together by closing the ocean that formed between them. For this to happen, the combined forces of slab pull and ridge push that had been driving dispersal must be overcome, requiring a reorganization of plate boundaries and subduction zones. In the alternative scenario, called extroversion, the continents continue outward and eventually collide on the far side of the globe by closing an exterior ocean. The relative strengths of ridge push and slab pull at various plate boundaries help determine which pathway a supercontinent cycle follows.
Why Venus Has No Plate Tectonics
Earth is the only rocky planet in the solar system with full-blown plate tectonics, and comparing it with Venus highlights just how critical slab pull is to keeping the system running. Venus is nearly the same size and composition as Earth, yet it shows no evidence of active subduction zones or spreading ridges.
One reason may be that Venus’ slabs are too buoyant to sink properly. On Earth, the transformation of basaltic crust into dense eclogite at depth is what gives slabs enough negative buoyancy to keep pulling. On Venus, higher surface temperatures mean the crust is hotter when it enters a hypothetical subduction zone, and the eclogite transformation may not proceed as thoroughly. Simulations of slab buoyancy show that Venus’ lighter slabs experience more resistance to subduction, making it harder for a self-sustaining plate tectonic system to develop.
This comparison underscores that plate tectonics is not just about having the right forces in principle. Earth’s particular combination of surface temperature, crustal composition, and mantle conditions allows slab pull to become strong enough to sustain itself. Change any of those variables, and the whole system may fail to get off the ground. Venus may be the clearest example of a planet where the slab-pull engine never quite caught.
How Models Capture These Forces Today
Modern geodynamic models simulate ridge push, slab pull, and mantle drag simultaneously in three dimensions, using seismic imaging of Earth’s interior to map out where slabs are, how dense they are, and how the mantle flows around them. These models prescribe weak plate boundaries using observed geometries and test how well the predicted plate motions match what GPS and seafloor measurements actually show.
Getting the plate boundary geometry right turns out to be as important as getting the forces right. High-resolution mantle flow models reveal that the precise shapes and locations of plate boundaries have a major effect on how slab pull forces translate into surface plate motion. A small change in where a subduction zone curves or where a transform fault offsets a ridge can alter the predicted velocity of a plate by a significant amount. The forces themselves may be relatively well understood, but the geometry through which they act is still being refined with every new seismic survey and GPS campaign.
Earth’s tectonic engine, in short, runs on gravity. Hot rock rises at ridges, cold rock sinks at trenches, and the plates are caught in between. Ridge push gives them a nudge; slab pull gives them a yank. Everything else, the mountain ranges, the earthquakes, the volcanic arcs, the opening and closing of oceans, follows from that basic gravitational tug of war playing out over hundreds of millions of years.