How Far Does a Bullet Travel Underwater?

Every action movie has the same scene: the hero dives underwater, incoming rounds slow to a harmless drift, and the villain empties a magazine into the pool for nothing. It is one of the rare moments where Hollywood lands closer to the truth than most people expect — and the reasons are worth writing down as equations, because they are more counterintuitive than "water is thick."

Water stops bullets fast. It stops rifle bullets faster than pistol bullets. And the projectile shape that wins in air is the wrong shape underwater. Here is the physics, with the math to back it.

1. The force: drag in a dense medium

A projectile moving through a fluid feels a retarding force set by the fluid's density, the projectile's frontal area, its drag coefficient, and the square of its speed:

Nothing in that expression changes between air and water except one term: the density ρ. Air is about 1.2 kg/m³; fresh water is about 1000 kg/m³. That is a factor of roughly 800. Cross the surface and the retarding force jumps by nearly three orders of magnitude, instantly, for the same bullet at the same speed.

2. The decay: how velocity dies with depth

To turn that force into a distance, write Newton's second law along the path and convert time to depth using v = dx/dt:

The v² on the right cancels one power of v on the left, leaving a simple separable equation. Integrating gives an exponential decay of velocity with depth:

The single parameter that governs everything is λ, the depth over which velocity falls to 1/e (about 37%). It rewards mass and punishes frontal area. A heavy, small-diameter bullet coasts further; a light, wide, or tumbling one dies quickly. Rewrite it as λ = 2·(m/A) / (ρ·C_d) and the ratio m/A is sectional density — so λ is simply SD divided by the resistance of the medium. The same property that governs long-range flight in air governs how far a bullet pushes through water.

Plug in three common loads. Frontal area is A = πd²/4; the drag coefficient C_d below is a nominal point-forward value, and the final row shows what happens when a bullet yaws sideways and presents its length instead of its nose:

3. Why the rifle round loses

The chart contains the whole counterintuitive result. In air, the 5.56 outranges both pistol rounds by a mile. Underwater, the stable 5.56 has the longest decay length (λ = 1.05 m) — but it never gets to use it.

Two things sink the rifle round. First, drag scales with v², so at 3,000 fps the retarding force is enormous — many times what an 1,100 fps pistol bullet feels. Second, that force acts on a long, light, fast projectile that is only marginally stable to begin with, so it yaws almost immediately. The instant it turns sideways, its presented area jumps from the nose (0.26 cm²) to the full side profile (about 1.08 cm²), and its drag coefficient roughly triples. Both terms in the denominator of λ blow up at once:

The decay length collapses by a factor of fourteen. The fast round burns itself out in the first few inches; the slow, heavy .45 plods furthest of the conventional loads. It is the tortoise and the hare, written in a drag equation.

4. Energy dies twice as fast

Wounding potential tracks kinetic energy, not velocity, and energy goes as v². Squaring the decay law doubles the exponent:

So the energy curve falls off far faster than the velocity curve. A round can still be physically moving while carrying almost none of the energy it needs to do damage — which is exactly why "the bullet reached me" and "the bullet could hurt me" are different questions underwater.

5. Cavitation: why nose shape flips

There is a second regime the simple drag model hides. Whether a projectile can outrun water at all depends on the cavitation number — the ratio of the pressure holding water together to the dynamic pressure the projectile is generating:

At bullet speeds the denominator is gigantic, so σ is tiny — of order 10⁻³ or smaller. That is deep in the supercavitation regime, where a projectile can wrap itself in a bubble of water vapor and travel through gas instead of liquid. But a bubble only forms and stays attached if the nose sheds flow cleanly. A flat or truncated nose — a cavitator — does this; a pointed spitzer nose, optimized for low drag in air, does not. It drags, destabilizes, and yaws. Underwater, the aerodynamic winner becomes the loser.

6. What the Navy already measured

None of this is new. NOLTR 70-174, "The Performance of Small Arms Ammunition When Fired Into Water" (U.S. Naval Ordnance Laboratory), documented exactly this — the program grew out of Vietnam-era questions about sentries firing at swimmers and submerged charges. Their conclusion matched the math above: a standard ogival, spin-stabilized bullet that flies well in air is unstable in water, and high-velocity rounds lose lethality within a very short distance. It is a good reminder of why leaning on primary sources beats trusting numbers repeated secondhand.

7. So — could you hide underwater?

Mostly, yes. A few feet of water between you and a shooter firing from the surface will defang most rounds, and counterintuitively you are better protected against a rifle than a pistol. Two caveats the equations don't show:

Angle. A bullet striking at a shallow angle can ricochet off the surface entirely, like a skipped stone. Steep, near-vertical shots penetrate; grazing shots skip.

Muzzle proximity. All of the above assumes the shot originates in air and crosses into water. Being close to the muzzle of a gun fired underwater is a different problem with a different answer.

8. The guns built to beat the physics

Militaries that genuinely need to shoot underwater discard the normal bullet entirely. The Soviet-era APS fires 5.66 mm steel darts about 120 mm long — flechettes tuned to supercavitate — effective to roughly 10–30 m depending on depth. Its successor, the ADS, fires supercavitating rounds and works on land too. Several companies now load supercavitating projectiles in conventional calibers. Every one of them throws away the spitzer nose and the light, fast bullet — the two things the drag equation and the cavitation number say will fail.

9. Calibrated numbers from the engine

The exponential model above is the shape of the physics with a constant drag coefficient. BallisticEngine's full solver refines it: it uses a Poncelet resistance law (a strength term plus the v² term, so the projectile stops at a finite depth rather than coasting forever), treats water as soft tissue so expansion applies, and accounts for the yaw and fragmentation that wreck the rifle round. Run the three loads and drop the calibrated output here:

Notice the ordering matches the physics: the slow, heavy .45 pushes furthest (122 in), the yaw-prone 5.56 far less (39 in), the 9mm least of all. But the 9mm's 13-inch result is dominated by expansion, not drag — the calibrated engine opens the hollow-point (RMF 0.60), roughly tripling its frontal area and crushing its λ well below what the constant-C_d table in section 2 predicts. That is the gap between the simplified model and the full solver: section 2 shows the shape of drag alone; a real JHP adds a second loss channel the exponential curve never sees. If the 9mm's cavity fails to open in water — an open question, since water lacks gel's fibrous structure — it would behave like an FMJ and push considerably deeper. Read the JHP row as the maximum-expansion case, not a guaranteed one.

* Penetration shown is depth to v ≈ 0 (projectile stops). This is not the same as "effective slant range" reported in NOLTR 70-174, which uses a ~300 fps lethality threshold — those distances are shorter.
† JHP expansion parameters are calibrated against ballistic gel; whether a hollow-point reliably expands in water is unresolved (see paragraph above).

The great equalizer

Water is the great equalizer of terminal ballistics. The drag equation only cares about density, area, and the square of speed — and it collects its toll in feet, not yards. Sectional density, velocity, and nose shape all still matter underwater; they just matter differently once the medium gets dense, which is exactly the kind of shift a penetration model exists to show.

Related reading: Sectional density: why it's not the whole story →

Frequently asked

Can a bullet kill you underwater?

Close to the muzzle, yes. But a bullet fired from above the surface loses lethal energy within a few feet of water — and rifle rounds lose it faster than pistol rounds.

How deep do you have to be to be safe from bullets?

For most surface-fired threats, a few feet of water is enough to stop or de-energize the round, though it depends heavily on caliber and shot angle. Deeper is always safer, and there is no single guaranteed number.

Do bullets travel farther underwater than in air?

No — dramatically shorter. Water is about 800 times denser than air, so drag rises by nearly three orders of magnitude and velocity decays exponentially with depth.

Why do rifle bullets stop faster than pistol bullets underwater?

Drag rises with the square of velocity, so a high-velocity rifle round meets far more resistance. That resistance destabilizes it into a yaw, which multiplies its presented area and collapses its penetration depth to a fraction of a stable round.

Can you fire a gun underwater?

Most modern firearms will physically fire underwater, but conventional ammunition is nearly useless past a very short range. Militaries use purpose-built underwater weapons that fire supercavitating darts instead.

Charts use a simplified quadratic-drag model, v(x) = v₀·e^(−x/λ) with λ = 2m/(ρC_dA), and the nominal inputs listed in section 2. Drag coefficients are representative, not measured; the yawed-5.56 case assumes a 90° presented profile. BallisticEngine's calibrated solver (Poncelet resistance, soft-tissue water model, yaw and fragmentation) produces the authoritative per-load numbers — the section 9 figures are model estimates, not lab measurements. Real results vary with projectile construction, entry angle, and depth.