In San Francisco and Washington, subway stations and platforms are under constant surveillance by closed-circuit television cameras. New Jersey Transit uses computer software that automatically alerts the police when an unattended package shows up on video monitors. The light-rail system in Houston plans to enable its onboard security cameras to transmit live images, wirelessly, to police cruisers.
In New York City, which has the largest mass transit system in the United States and is considered the country's most likely target for a terrorist attack, use of cameras is sporadic and not well coordinated. The Metropolitan Transportation Authority uses 5,723 cameras throughout its network, but they are not distributed uniformly; there are none, for instance, in close to half of the city's 468 subway stations.
Since the 1970's, transit agencies across the world have tried, first as a deterrent to combat vandalism, then with greater urgency after terror threats, to prevent and investigate crimes using electronic surveillance. In the United States, these efforts have produced a hodgepodge of systems, some advanced and others barely existent.
The need for, and usefulness of, surveillance was given fresh significance after the bombings in London last week, and in particular with the announcement on Tuesday that cameras had evidently helped to identify likely suspects, including at least one presumed suicide bomber.
But while cameras were evidently critical in this investigative breakthrough, the London attack also underscored their limitations: the cameras, of course, could not and did not prevent the explosions. Similarly, video cameras captured some of the movements of the Sept. 11 hijackers, but those images chiefly served, in the aftermath of the attacks, as a kind of eerie visual diary.
Officials in London, which has one of the most extensive surveillance networks in the world, believed that by making cameras omnipresent they would deter bombers, particularly the Irish Republican Army.
''It obviously does not deter suicide bombers,'' said Brian M. Jenkins, a consultant at the RAND Corporation, a policy research organization, ''but it did have a measurable effect on the I.R.A.'s campaign.''
But the proliferation of cameras came with practical problems, too.
''Unless you have thousands of people watching thousands of cameras, you don't have the same effect as with actual police presence,'' Mr. Jenkins said.
In fact, the whole notion of using cameras to fight crime, including terrorism, inside the world's transportation networks is fraught with financial and technological challenges.
The traditional model of surveillance cameras -- a bank of monitors with fuzzy black-and-white images, under a person's inattentive gaze -- is being replaced by digital video recorders that transmit, record and store data over fiber-optic or wireless networks. But the new software and wireless enhancements mean that a single custom-built camera can cost over $10,000 to acquire, install and maintain.
Nonetheless, the use of video surveillance has already become a major thrust of transit security programs in the United States.
''The technology now allows a very quick replay and return to specific points in time, and much better quality in terms of the resolution of images,'' said Greg Hull, director of operations safety and security at the American Public Transportation Association.