How Zebra and Pelican Crossings Work and What Flashing Lights Mean

The world's first traffic signal, put up outside the Houses of Parliament in December 1868, ran on gas, and in January 1869 it exploded, injuring the police constable who operated it. After that, work on signalled crossings stopped for about fifty years. What replaced it was more than a century of experimentation, from black-and-white stripes tested in Slough to flashing greens that mean three different things in three different places. Where a pedestrian may cross, who must stop, and what a light means when it flashes all depend on a set of habits that were settled country by country.

How does a pedestrian crossing work?

A pedestrian crossing, known as a crosswalk in American and Canadian English, is a place designed for people to get across a road, street or avenue. The Vienna and Geneva Conventions both use the term. Marked crossings often sit at intersections, but they are also placed on busy stretches that would be unsafe to cross unaided because of the number of vehicles, their speed or the width of the road. They are common where many people cross, such as shopping areas, or where vulnerable users like school children cross regularly.

The rules differ. In some areas the driver must wait until the pedestrian has got past the midpoint of the crossing. In others, jaywalking laws restrict crossing anywhere other than marked facilities, though even where such laws exist, an unmarked crossing is generally assumed at every intersection unless a sign says otherwise.

The main split is between signal and non-signal crossings. Signals clearly separate the time for pedestrians from the time for vehicles. Crossings without signals generally help pedestrians and often give them priority, depending on locality. A pelican crossing uses signals to gather pedestrians where drivers can see them and cross them most safely, while a zebra crossing is uncontrolled and better suited to lower traffic flows. Some crossings that look like ordinary ones are created mainly to calm traffic, especially when combined with pedestrian priority, refuge islands or raised surfaces.

Invented to get Members of Parliament across a London street in time to vote, and it injured the constable who ran it.

Where did pedestrian crossings and the zebra stripe come from?

The 1868 signal at Bridge Street was meant to let Members of Parliament cross to the Parliamentary Estate in time to vote. Its inventor, railway engineer John Peake Knight, built it from three semaphore arms topped with a gas lantern showing green and red at night, and a constable turned a handle at the foot of the pole to raise and lower the arms.

Cars changed the problem. In 1911 a reader of The Times wrote that pedestrians had lost their old margin of safety, now that vehicles were faster and drivers seemed unwilling to slow until the last moment. By 1949 Britain's Ministry of Transport had put 1,000 experimental crossings on its roads during a Pedestrian Safety Week, and on 31 October 1951 a striped crossing was marked at Slough, west of London, described as the first pedestrian crossing ever marked. The black and white stripes were nicknamed "zebra crossings". Nobody can say with certainty where the name came from, though the likeness to zebra fur is the usual explanation, and it is believed the politician James Callaghan first used the term.

The stripes were designed to be noticed. The first zebra installation had evenly spaced white bars running parallel to the traffic, generally 24 inches wide with gaps of the same size, on the theory that the full width of the crossing would be conspicuous to drivers and pedestrians alike. An early test by the Department of Scientific and Industrial Research in 1948, before any public education, lifted the proportion using the right spot among men from 45% to 53%, among women from 63% to 69%, and among children from 69% to 81%. Transverse, longitudinal and diagonal stripes and checkerboards were also tried, and the zebra won. The Ministry made it the standard in October 1951.

The results looked good. Pedestrian casualties in Great Britain were lower in 1952 and 1953 than in the years before. Dr W. H. Glanville of the Road Research Laboratory reported a 7% fall in pedestrian accidents, with towns that had crossings improving by 8% against 2% in rural districts, and a 2% rise in Northern Ireland, where nothing had changed. Norfolk, Virginia, one of the first US cities to test the zebra pattern, in 1953, had averaged 4 pedestrians injured a year at plain crosswalks and recorded none in the three years after.

What types of pedestrian crossing are there?

Unmarked crossings are presumed at intersections in some countries, including the US, except where crossing is expressly prohibited. A pedestrian refuge is an uncontrolled crossing with two dropped kerbs and a central island. It lets people cross one direction of traffic at a time and reduces pedestrian accidents by around 40%, and it narrows the road, which slows vehicles and discourages overtaking. But it may not give pedestrians priority, so waits can be longer than at a controlled crossing, and the pinch points it creates can be dangerous for cyclists. A courtesy crossing has coloured surfacing or a similar hint that people may cross, aimed at getting drivers to give way out of courtesy rather than obligation. Stripes in the paving and visual narrowing of the road both increase that courtesy.

Signal crossings may need a button to trigger them, and may add audible or tactile signals for people with poor sight. Many cities fit countdown timers. Where foot traffic is very heavy, systems with flashing lights embedded in the pavement may warn drivers, or an exclusive pedestrian phase, a Barnes Dance, stops vehicles in all directions at once. A pedestrian scramble does something similar: it shows red to every vehicle for a period so people can cross in any direction, diagonally included. A crossing can be single-stage, crossing the whole junction in one go, or multi-stage, with an island and a second signal. Toronto mandates single-stage crossings unless they would cause capacity problems from a longer signal cycle.

At very busy junctions or limited-access roads, footbridges and tunnels may replace crosswalks, though they are expensive and so spaced far apart, and ramps, stairs and lifts put people off. Pedestrians often prefer to cross at ground level instead.

Deciding where a crossing goes

Some authorities lean on a formula. PV2, developed in the United Kingdom in 1987 and later adopted in India and Iran, multiplies the pedestrian flow by the square of the vehicle flow in the peak hour. With 500 pedestrians and 600 vehicles an hour, it gives a figure that calls for a pelican crossing. The US Manual on Uniform Traffic Control Devices says crosswalk markings should not be used indiscriminately and warns against them, without extra engineering measures, on high-traffic routes with a limit above 64 km/h (40 mph). Critics say such guidance lets engineers focused on motorists avoid safety measures. Some activists respond with guerrilla crosswalks, painted where official requests have been ignored; in Los Angeles these are often removed, a few are tolerated, and police have occasionally made arrests.

Where do traffic light sequences come from?

Traffic lights usually have three signals, red to stop, amber to signal a change and green to go, in that order from top to bottom or left to right. That much is international, but sequences and laws differ nationally and locally. Amber is yellow in North America and orange in Australia, and in South Africa, Zambia and Namibia the lights are called robots.

After the exploded gas lantern, the next advance was electric. Lester Wire, a Salt Lake City policeman, developed the first electric light in 1912, and it was installed in Cleveland in August 1914. In 1920 William Potts of the Detroit police created the first four-way, three-colour signal and the first to add amber, then installed automatic versions on 15 towers the next year. By 1922 timers were replacing officers, and New York City freed up everyone but 500 of the 6,000 officers on its traffic squad, a saving of $12,500,000. Garrett Morgan patented a manually operated three-way signal with moving arms in 1923. Computers arrived in the 1950s, with Denver as an early example in 1952, and in 1967 Toronto used more advanced machines to control 159 signals over telephone lines.

What the standards say

Chapter III of the 1968 Vienna Convention sets international standards, though not every state has ratified it. A three-colour head should have steady red, amber and green lights, stacked with red on top or side by side with red opposite the traffic flow, all circular or arrow-shaped. The United States is not a party. It follows the MUTCD, under which a signal head may have three, four or five aspects, and a single flashing amber or red can draw attention to a warning or stop sign.

The British sequence is red, red with amber, green, then amber, where red and amber means get ready but do not go and amber means stop unless it is unsafe. Japan calls the green light "ao", normally translated as blue, a leftover of how the language changed, and in 1973 officials decreed that the go light should be the bluest possible shade of green. Quebec and the Maritime provinces often hang lights horizontally, with a square red, a diamond amber and a circular green.

What do flashing lights mean?

The flashing signals carry the most local variation. In the United States and Canada, flashing yellow does not require a stop, but drivers should take care as opposing traffic may enter after stopping. It appears when signals malfunction or late at night. In Canada, a flashing yellow at a junction that would otherwise have signals means they have failed, and everyone must treat it as a four-way stop until repairs are made. All-way yellow signals are prohibited by US regulations, though all-way red flashing can supplement all-way stop control.

In the United Kingdom, an amber light precedes the flashing red lights used outside fire and ambulance stations, at level crossings and airfields and at lifting or swing bridges, though not at London's Tower Bridge, which uses ordinary red lights. At some stations flashing blue lights tell the emergency driver that the road lights are working. Pelican crossings, with their flashing amber phase, are being gradually replaced by puffin or Pedex crossings, because the flashing phase confused pedestrians and drivers alike. Twin flashing ambers also appear at cattle and army tank crossings.

Flashing green is the most contested. Across most of Canada it means an advanced green, allowing a left turn across oncoming traffic without yielding. That choice dates from when a green arrow in Ontario meant everyone had to turn, and an advanced green was also easier to add to existing three-lamp signals. In British Columbia it means something else entirely, a signal controlled by pedestrians and cyclists that rests on green, which can confuse visitors, and Ontario is phasing out flashing greens in favour of arrows. The 2009 US MUTCD prohibits flashing any green, which also breaches the Vienna Convention, as does the use in Austria, Croatia and Estonia of flashing green as advance notice of amber.

Arrows and yellow traps

Across most of the United States, a protected turn is shown by a steady green arrow 300 mm (12 inches) across, followed by a yellow clearance interval of not less than three seconds. The flashing yellow arrow, a permissive left-turn signal in which drivers must yield to oncoming vehicles and pedestrians, is meant to prevent the yellow trap. That trap arises when the signal turns yellow and a turning driver assumes oncoming traffic will stop, and the US bans sequences that cause it or requires signs warning of extended green. The Minnesota Department of Transport says the flashing yellow arrow reduces delays and increases safety, and a national research programme found it safer than a circular green. Maryland, though, in its 2011 manual preferred a flashing red arrow.

Elsewhere, arrows work differently. A UK filter arrow is a green arrow shown while the main signal is still red, meaning that movement conflicts with no other. Quebec sometimes shows a lone green straight arrow for five to nine seconds so that pedestrians can get into the road first. In New Zealand a red arrow can hold turning traffic while a pedestrian crossing on the side road has its green, a method becoming common in Australian states. Australia and New Zealand also use doghouse signals, with six signals including a separate red arrow.

Countdowns, pedestrian signals and the dilemma zone

Pedestrian signals usually have two lights, a walking figure in green or white and a "don't walk" figure or hand in red or orange. Countdown lights are popular in Vietnam and China, with a number that reaches zero as the main light changes. The Philippines used them for vehicles and pedestrians too, but since 2025 Metro Manila has phased them out in favour of volume-based signalling, in which the green blinks five times before turning yellow.

South Korea shows how strict a rule can be. Its law does not recognise a yellow-light dilemma zone: when amber shows, traffic may not cross the stop line, even if it cannot stop safely. In May 2024 the Supreme Court of Korea reaffirmed this in a case in which the driver was doing 62 km/h in a 40 km/h street. Critics call it unrealistic and say sudden braking causes multiple collisions.