Shark Attacks Kill 10 People a Year — Here's Why That Stat Is Useless

Ten Deaths. Eight Billion People. Do the Math.

Every year, somewhere between four and ten people are killed by sharks worldwide. The International Shark Attack File (ISAF), maintained by the Florida Museum of Natural History, is the most authoritative dataset on the subject: it logged ten unprovoked shark fatalities in 2023, five in 2022, nine in 2021, and four in 2024. The ten-year average sits at about six. The number is strikingly stable, and as a global statistic, almost entirely without meaning.

Take the 2023 count. Ten deaths divided by eight billion people is 0.00000000125, roughly 1-in-800-million. Your odds of being killed by a shark in a given year, as a random member of the global population, are essentially zero. Technically accurate. Deeply misleading.

The Paradox of the “Safe” Statistic

The ten-deaths figure has done strange cultural work. It appears in rebuttals to shark-scare media coverage, and those rebuttals are, in an important sense, correct. Jaws-style hysteria about lurking predators has distorted public perception. Sharks do not hunt humans. Most encounters end without incident. From a policy perspective, culling programs have repeatedly been shown to be ineffective and ecologically damaging.

But the corrective went too far. The response to irrational fear was to produce a number so small it erased the risk entirely. “More people die from vending machines than sharks” became a standard line, rhetorically satisfying but analytically empty. Billions of people interact with vending machines. Very few regularly swim in open ocean water near shark populations.

The ten-death figure is fine as a population-level count. It fails the moment anyone treats it as a personal risk estimate.

Who Actually Swims in the Ocean?

The global population is approximately eight billion. Of those, a significant fraction lives nowhere near an ocean. Many who live near the ocean do not swim regularly. Many who swim do so in pools, calm bays, or waters with low shark populations.

The relevant denominator is not eight billion. Nobody counts ocean swimmers directly, so this site works with an estimate: roughly 300 million person-hours spent in the ocean each year worldwide. That figure is a back-calculation, not a survey result. Around 60 million Americans swim in coastal waters in a typical year, the United States accounts for roughly a third of recorded shark bites, and a few hours per swimmer per year gets you to about that total. Treat it as an order-of-magnitude anchor.

The geographic concentration reinforces the point: unprovoked attacks cluster heavily in a handful of regions. Florida accounts for more US attacks than all other states combined. The “shark attack triangle” of Florida, Australia (particularly Western Australia and New South Wales), and South Africa accounts for the large majority of global incidents year after year.

A person who swims regularly in the surf off Daytona Beach, dives in South Africa’s Aliwal Shoal, or surfs the breaks off Western Australia is not facing the same probability as someone who wades into the Mediterranean once a decade on vacation. The global average blends these wildly different populations together and tells you nothing useful about your personal exposure.

The Real Math

Working from ISAF data: roughly 70 unprovoked shark bites occur globally per year (the ten-year average), of which about six are fatal. Divided among an estimated 300 million ocean swim-hours annually, that yields a bite rate of about 1-in-4-million per swim-hour and a fatality rate of about 1-in-40-million per swim-hour. Attacks are not uniformly distributed, though. The rate climbs meaningfully in high-incidence zones.

The probability chain has three links. A shark must be in your vicinity during your swim, which depends on coastal region, time of day (dawn and dusk carry higher risk), water clarity, baitfish activity, and seal populations nearby. Given an encounter, the shark must investigate aggressively. ISAF consistently reports that most encounters do not result in bites; the calculator uses a 3% bite rate given a true encounter, a working assumption rather than a measured figure. Given a bite, the attack must be fatal. The ISAF record for the past decade shows roughly a 10% fatality rate for unprovoked bites (six deaths among about 70 bites a year), reflecting that medical response and the “bite and release” behavior of most sharks keeps deaths well below total attacks.

Multiply those three probabilities across hours in the water and days per year, and the per-person annual risk becomes calculable. A regular surfer who spends 200 days a year in the ocean faces meaningfully different numbers than a tourist who swims for an hour on one beach holiday.

Your Actual Risk

The calculator below lets you model your specific exposure. The two variables that move the needle most are hours per day in the water and days per year you swim in the ocean. A once-a-year beach vacation produces a trivially small number. A professional surf instructor working daily in coastal waters produces something more sobering.

At the default setting of one hour in the water on 50 days a year, the calculator returns about 1 in 670,000. Push it to an hour every day of the year and it lands near 1 in 91,000.

Drowning Puts It in Perspective

The comparison that matters is drowning. The CDC reports roughly 4,000 unintentional drowning deaths per year in the United States, which across the whole population works out to about 1 in 82,000 per year. That is a population rate, not a swimmer rate; a swimmer-specific figure would be higher, but no agency publishes one.

Even against that diluted benchmark, sharks lose badly. At the calculator’s default exposure, shark fatality risk (about 1 in 670,000) is roughly eight times lower than the general-population drowning rate. Only a daily ocean swimmer gets close to parity, and that swimmer’s own drowning risk is well above the population average too.

If you are a regular ocean swimmer, drowning is the water-entry risk you should be thinking about. Rip currents kill. Exhaustion kills. Cold water incapacitation kills. Sharks kill too, but far less often for the same water exposure.

The globally-averaged ten deaths per year is irrelevant to your decision-making. What matters is your specific exposure and whether you’ve honestly accounted for the much larger risks that share the same environment.

Car crashes kill 1-in-8,000 Americans per year. Lightning kills about 1-in-16,000,000. These are the benchmarks for contextualizing any other mortality risk. Shark attacks, even for regular ocean swimmers in high-incidence regions, fall below the car crash level. But not by the factor of millions that the global average implies.

Sources

• International Shark Attack File (ISAF), Florida Museum of Natural History. Yearly worldwide shark attack summaries, 2021–2024. floridamuseum.ufl.edu/shark-attacks. Unprovoked fatalities: 9 (2021), 5 (2022), 10 (2023), 4 (2024); ten-year average about 6 fatalities and 70 bites per year.
• Naylor, G. J. P. and Miguez, J. (2024). The ISAF 2023 Shark Attack Report. Florida Museum of Natural History.
• ISAF methodology note: “Unprovoked attacks” are defined as incidents where a shark attacks a live human in the shark’s natural habitat without human provocation. Provoked attacks (handling sharks, spearfishing, etc.) are excluded from the primary count.
• Ocean exposure denominator: site estimate of roughly 300 million person-hours in the ocean per year worldwide. Derived from the 2010 NOAA-cited figure of about 60 million Americans swimming in coastal waters, the US share of ISAF bites (roughly one third), and an assumed few hours per swimmer per year. Not an ISAF figure.
• Fatality rate: ISAF ten-year averages, about 6 deaths per 70 unprovoked bites (roughly 10%).
• US drowning statistics: Centers for Disease Control and Prevention, Drowning Facts, average of about 4,000 unintentional drowning deaths per year (2012–2021). cdc.gov/drowning
• US lightning deaths: National Weather Service, about 20 deaths per year in recent years.