When it happens, it usually makes banner headlines.
A 50-year-old woman was mauled to death last August by a great white shark off the California coast. Two months later, what was probably a tiger shark attacked a 13-year-old surfing champion off the coast of Hawaii and severed her left arm.
Both events were international news, but they were the exception.
For the third consecutive year, the number of shark attacks worldwide dropped, according to a study released Tuesday by George Burgess, director of the International Shark Attack File at the Florida Museum of Natural History.
Last year's total of 55 unprovoked attacks worldwide was down from the 63 attacks in 2002, the 68 in 2001 and the 79 in 2000.
Florida retained its title as shark attack capital of the world with 31 reports last year. But even that was somewhat lower than the 2000-02 average of about 33. There were 29 attacks in 2002, 34 in 2001 and 37 in 2000.
The highest number of attacks in Florida (14) occurred on the central East Coast in Volusia County, and surfers were the most frequent victims, followed by swimmers and waders, and divers and snorkelers.
The cause of the decline, Burgess said from his Gainesville office, may be the result of a long-term trend, such as the possibility there are fewer sharks and people in the water together. The number of shark attacks in any given year is directly related to the amount of time humans spend in the water as well as the number of sharks living there, Burgess said.
But it could be just as likely there are simply fewer sharks.
And that, he said, is troubling.
Shark populations have declined worldwide in recent years because of overfishing, Burgess said, and because most sharks don't reach sexual maturity until age 8 or older, they are especially vulnerable.
"In many ways, sharks grow somewhat like humans," Burgess said. "They can live to be 40 or 50 years old, they are slow to reach sexual maturity, have a limited number of young, and are slow growing.
"So when a population gets in decline among these animals, recovery time is very slow, even if you stop all fishing mortality. In the case of sharks along the East coast, because of overfishing, recovery will take probably 30 years even with strong fishery measures that are in place."
The study also showed that there were four deaths worldwide from shark attacks last year, compared with three in 2002, four in 2001 and 11 in 2000.
Still, the number is small when compared to deaths involving other animals. The average number of fatalities per year in the United States is greater for deer (130, mostly through vehicular collisions), dogs (18), snakes (15) and mountain lions (0.6), than it is for sharks (0.4).
But a shark attack is usually far more dramatic than a dog bite.
"When sharks bite people, they get a lot of attention. A disproportionate amount of attention," said Sonja Fordham, a shark fisheries specialist with the Ocean Conservancy in Washington, D.C.
The conservancy doesn't use shark attacks as a measure of shark population. "We tend to look toward scientific workshops convened by the government," Fordham said, "But they include people like George Burgess, and they are alarming enough.
"We already know our coastal sharks are overfished, and that some species are severely threatened. These are some of the most vulnerable animals in the ocean. Things are changing, but not fast enough."
So while the decline in attacks is encouraging for surfers, bathers and swimmers, it could be more unwelcome news for those who track shark populations.
"The bottom line," Burgess said, "is that sharks are paying a heavier consequence than are humans. Each year humans kill about 60-million sharks worldwide. But sharks kill between three and 10 people a year worldwide.
"You can make your own decision as to who is the winner and the loser."