What are the odds of drowning in a sinking car?
Evidence quality 4.5/5
Eight-dimension review score against the quality rubric . Each dimension scored 1–5.
- D1 Source grounding
- 4/5
- D2 Source authority
- 5/5
- D3 Arithmetic
- 5/5
- D4 Uncertainty
- 5/5
- D5 Scope
- 4/5
- D6 Prose
- 4/5
- D7 Perception honesty
- 4/5
- D8 Caveat completeness
- 5/5
● your factors — click this risk ▾ to reveal
- Your factors
≈ As likely as
Perceived
The sinking-car scenario is a fixture of nightmares and driver-anxiety threads: the vehicle goes off a bridge or into a canal, the water rises, the power windows fail, and there is no way out. It is dramatized often enough that many people have a rehearsed mental escape plan for it while having none for the far likelier ways a car kills. The dread is partly rational — submersion is genuinely lethal once it happens — but the event itself is rare, and the perceived probability sits far above the measured one.
Rough estimate: widely feared as a real possibility; the actual per-lifetime risk is on the order of 1 in 17,000
Source: editorial intuition, not polled
Actual
~350 US vehicle-submersion drowning deaths per year (5-11% of ~4,345 annual drownings; ~400/yr across North America)
US residents; derived from CDC drowning totals × submersion share (Giesbrecht & McDonald)
Show derivation
Giesbrecht & McDonald (2010) estimate ~400 vehicle-submersion deaths per year across North America, accounting for 5-11% of all drownings. The CDC reports an average of 4,345 unintentional drowning deaths per year in the US (2018-2021); the 5-11% submersion share implies ~217-478 US submersion deaths per year, and the US accounts for the large majority of North America's drownings, so ~350/yr is a reasonable central US figure. Spread across ~335 million residents that is ~1.05e-6 per person per year; over a 59-year adult horizon, 1-(1-1.05e-6)^59 ≈ 6.2e-5, or roughly 1 in 16,700. The uncertainty band spans the 5% submersion share (low) to a generous allowance above the 11% share (high), and also absorbs Giesbrecht's observation that submersion deaths are systematically *under*-counted — many never enter crash statistics because no accident report is filed when the outcome is a drowning. This is a population-average figure; driving into floodwater and failure to exit during the float phase move individual risk by close to an order of magnitude (see factors).
Caveats: Submersion deaths are under-counted. As Giesbrecht & McDonald note, when the out…
Submersion deaths are under-counted. As Giesbrecht & McDonald note, when the outcome is a drowning, local police often do not file a crash report, and a vehicle pulled from water by a fire department may never be coded as a submersion crash — so the true figure likely sits toward the upper end of the range. The US central estimate (~350/yr) is derived by apportioning the ~400/yr North America figure and cross-checking against the CDC drowning total times the 5-11% submersion share; it is not a directly published US count. The escape-window and floodwater multipliers are grounded in the NWS and Giesbrecht findings but are directional rather than precise relative risks. The dramatized scenario (off a bridge into deep water) is real but a minority of cases; the statistically dominant one is mundane — a driver entering a flooded road that looks shallower than it is.
Related risks
Other risks on similar themes — for exploring related fears.
Boating drowning
What are the odds of drowning after falling overboard from a recreational boat or yacht?
Child without restraint
How much more likely is a child to die in a car crash without an appropriate child restraint?
Pick challenger
The sinking car is one of the more rehearsed driving fears — people who have never thought about how to survive a rollover can describe the plan for a submersion: seatbelt off, break the window, get out before it goes under. The dread is not misplaced about consequences. Submersion is highly lethal once it happens, and researchers Gordon Giesbrecht and Gerren McDonald estimate roughly 400 deaths a year across North America, accounting for 5 to 11 percent of all drownings. Against the CDC’s ~4,345 annual US drownings, that share implies on the order of 350 US submersion deaths a year — a lifetime probability near 1 in 16,700.
The dramatized scenario — a car off a bridge into deep water — is real but a minority of cases. The statistically dominant one is far more ordinary: a driver entering a flooded roadway that looks shallower than it is. The National Weather Service reports that over half of all flood-related drownings happen when a vehicle is driven into floodwater, and that just twelve inches of moving water can carry away most cars. The most consequential decision is made before any escape is needed: whether to drive into the water at all. “Turn around, don’t drown” is a slogan, but it maps directly onto where the deaths actually occur.
When submersion does happen, survival is a function of timing. Giesbrecht’s work shows vehicles float for roughly fifteen to sixty seconds before water reaches the windows, and that float phase is the survivable escape window — belts off, a window down or broken, occupants out. After the cabin fills, water pressure makes doors very hard to open and electric systems may fail, which is why most preventable deaths come from attempting escape too late rather than from being physically trapped from the start. The figure is also probably an undercount: when the outcome is a drowning, the incident often never enters crash statistics at all. The fear is calibrated to the wrong picture — it imagines the bridge, when the danger is the flooded street.
Related tidbits
About 1 in 16,700 lifetime for a US resident, roughly 350 vehicle-submersion drowning deaths a year. It is a vivid, widely feared scenario, but it accounts for only 5-11% of US drownings.
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
-
[1] Giesbrecht GG, McDonald GK — Aviation, Space, and Environmental Medicine, 2010 Aug;81(8):779-84 — My car is sinking: automobile submersion, lessons in vehicle escape
My car is sinking: automobile submersion, lessons in vehicle escape- Statistic
~400 vehicle-submersion deaths per year in North America, 5-11% of all drownings; vehicles float ~15-60s before windows submerge; total time to submersion 37-150s depending on body integrity- Excerpt
“"In North America, approximately 400 individuals per year die in submersed vehicles, accounting for 5-11% of all drownings." ”
- Source data from
- 2010-08-01
- Accessed
- 2026-06-11 · archived copy
- Calculation
- Primary source for the submersion-specific count and the escape window. The ~400/yr North America figure × the US share of North American drownings gives the ~350/yr US central estimate. The float-phase finding (vehicles stay buoyant for roughly 15-60 seconds before water reaches the windows) underpins the escape-window multiplier: this is the survivable period.
-
[2] U.S. Centers for Disease Control and Prevention (CDC) — Drowning Data and Statistics
Drowning Data and Statistics- Statistic
An average of 4,345 unintentional drowning deaths per year in the US, 2018-2021- Excerpt
“"An average of 4,345 unintentional drowning deaths occurred each year from 2018-2021." ”
- Source data from
- 2024-05-01
- Accessed
- 2026-06-11 · archived copy
- Calculation
- Supplies the US drowning denominator. Applying the 5-11% vehicle-submersion share from Giesbrecht & McDonald to 4,345 annual drownings gives ~217-478 US submersion deaths per year, bracketing the ~350 central estimate and setting the uncertainty range.
-
[3] National Weather Service (NOAA / NWS) — Turn Around Don't Drown — Flood Safety
Turn Around Don't Drown — Flood SafetySee all 2 Likelier entries citing this source →
- Statistic
Over half of flood-related drownings occur when a vehicle is driven into floodwater; ~12 inches of moving water can carry away most cars- Excerpt
“"Over half of all flood-related drownings occur when a vehicle is driven into hazardous flood water. It takes just 12 inches of rushing water to carry away most cars and just 2 feet of rushing water can carry away SUVs and trucks." ”
- Source data from
- 2024-01-01
- Accessed
- 2026-06-11 · archived copy
- Calculation
- Establishes that driving into floodwater is the dominant submersion mechanism, not bridges and canals. Basis for the floodwater multiplier: the single largest controllable risk factor is the choice to enter a flooded roadway.