An action potential propagates along an axon by triggering a chain reaction of electrical changes in the membrane, where each segment of the axon depolarizes just enough to push the next segment past its threshold voltage. The signal doesn’t travel like electricity through a wire. Instead, it’s regenerated fresh at each point along the way, which is why it arrives at the far end of the axon at full strength.
Resting State and Threshold
Before any signal fires, a neuron’s membrane sits at a resting voltage of roughly -70 to -80 millivolts, meaning the inside of the cell is negatively charged relative to the outside. This voltage exists because the membrane is far more permeable to potassium than sodium at rest, and potassium ions leak outward, leaving behind a net negative charge. The resting potential for potassium alone would be about -90 mV, but small amounts of sodium leaking in pull the actual resting value slightly more positive.
For an action potential to begin, something has to push the membrane voltage positive enough to cross a critical threshold. This threshold is the voltage at which voltage-gated sodium channels snap open. Below it, small voltage changes fade out passively. Above it, a self-amplifying cascade begins.
The Rising Phase: Sodium Rushes In
Voltage-gated sodium channels have internal sensors that respond to changes in electrical charge across the membrane. At resting potential, those sensors are pulled inward, mechanically pinching the channel’s pore shut. When the voltage rises past threshold, the sensors swing outward, releasing that mechanical block and opening the pore. Sodium ions flood into the cell because there’s far more sodium outside (about 140 millimolar) than inside (about 14 millimolar), and the negative interior pulls them in electrochemically. This inward sodium current drives the membrane voltage sharply positive, toward sodium’s equilibrium potential of around +60 mV.
This is an all-or-none event. If threshold is reached, the full action potential fires. If it isn’t, nothing happens. There’s no half-strength action potential.
The Falling Phase: Potassium Flows Out
Two things terminate the spike almost as quickly as it began. First, the sodium channels inactivate. After opening, they automatically transition into a blocked state within a fraction of a millisecond, shutting off the sodium current even while the membrane is still depolarized. Second, voltage-gated potassium channels open. These channels respond to the same depolarization that opened the sodium channels, but they’re slower to activate. By the time they fully open, the sodium channels are already closing. Potassium ions then flow out of the cell, driven by both the concentration gradient and the now-positive interior voltage. This outward potassium current rapidly pulls the membrane back toward its resting potential.
The potassium channels are slow to close as well, so they briefly drive the membrane slightly more negative than the normal resting value, a dip called hyperpolarization. The voltage then gradually returns to its baseline as these channels finish closing.
How the Signal Moves Forward
Here’s the key to propagation: when one patch of membrane depolarizes to +60 mV, that positive charge spreads passively to the neighboring stretch of membrane, like a wave of warmth moving along a surface. This passive spread, called electrotonic conduction, pushes the adjacent region past its threshold voltage. That region then fires its own full action potential, which in turn depolarizes the next region, and so on down the axon. Each segment regenerates the signal from scratch, so it never weakens with distance.
In unmyelinated axons, this process happens continuously along every micrometer of membrane. That’s relatively slow. Unmyelinated C-fibers, the thin nerve fibers that carry dull pain and temperature signals, conduct at roughly 1 meter per second.
Why the Signal Only Moves in One Direction
When a patch of membrane fires an action potential, the sodium channels in that region enter their inactivated state. During this absolute refractory period, no stimulus of any strength can reopen them. They need time to reset. This means the backward-facing membrane, the region that just fired, is temporarily unable to depolarize again. The only region capable of reaching threshold is the forward-facing membrane that hasn’t yet fired.
After the sodium channels recover from inactivation, there’s a brief relative refractory period during which the lingering outward potassium current makes it harder (but not impossible) to trigger another spike. Only an unusually strong stimulus could fire an action potential during this window. Together, these refractory periods ensure signals travel in one direction: from the cell body toward the axon terminal.
Saltatory Conduction in Myelinated Axons
Many axons in the brain and body are wrapped in myelin, a fatty sheath formed by supporting cells. Myelin doesn’t cover the axon continuously. It’s interrupted at regular intervals by small gaps called nodes of Ranvier, where the axon membrane is exposed and packed with voltage-gated sodium channels.
Myelin dramatically speeds up propagation by lowering the membrane’s electrical capacitance, meaning the membrane stores less charge and voltage changes travel farther and faster through the insulated segments. When an action potential fires at one node, the resulting depolarization spreads passively through the myelinated internode and arrives at the next node still strong enough to cross threshold. A fresh action potential then fires at that node, and the process repeats. The signal effectively jumps from node to node, a process called saltatory conduction (from the Latin “saltare,” to jump).
Research on myelinated fibers has shown that conduction speed is largely optimized with about 16 layers of myelin wrapping. Adding more layers beyond that provides only modest additional gains. The primary benefit comes from reducing capacitance rather than simply insulating the axon against current leakage. Myelinated A-delta fibers, which carry sharp “first” pain, conduct at roughly 15 meters per second, about 15 times faster than unmyelinated C-fibers. The largest myelinated motor and sensory fibers conduct even faster, with conduction velocity scaling proportionally with axon diameter across a range that varies more than 100-fold.
Restoring the Ion Gradients
Each individual action potential moves only a tiny number of ions across the membrane, so the concentration gradients don’t change meaningfully after a single spike. But over thousands of action potentials, sodium would gradually accumulate inside the cell and potassium would be depleted. The sodium-potassium pump, an enzyme embedded in the membrane, prevents this by continuously shuttling sodium back out and potassium back in, using energy from ATP. This pump doesn’t generate the action potential or drive its propagation directly. Its role is maintenance: keeping the ion gradients intact so the next action potential has the same electrochemical driving forces as the last.
This is why nerve signaling has a real metabolic cost. The brain, despite being only about 2% of body weight, consumes roughly 20% of the body’s energy at rest, and a significant fraction of that energy goes to running sodium-potassium pumps.