U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-RD-99-089
Date: December 1999

Research, Development, and Implementation of Pedestrian Safety Facilities in the United Kingdom


5. Pedestrian Crossings without Signal Control (Crosswalks)

5.1 Zebra crossings

Zebra crossings were introduced in their current design in the 1950s and are still in widespread use in the United Kingdom. They are indicated by black and white bands painted on the surface of the carriageway. Since 1971 zigzag lines have been painted upstream and downstream of Zebra crossings to alert drivers to the crossing and prohibit overtaking and parking close to the crossing. There may or may not be a pedestrian refuge at the mid point on the crossing. (See figure 9.)

Drivers are required, under the Highway Code, to stop for pedestrians on Zebra crossings. Legally, pedestrians have to establish precedence by standing on the crossing. UK drivers often also stop when they see a pedestrian waiting to cross — something that does not appear to happen at Zebra crossings in other countries.

Over the past 10 years, many Zebra crossings have been replaced by Pelican crossings (signal control), and new crossings tend to be Pelicans rather than Zebras. There is now an estimated 9,000-10,000 Zebra crossings in the United Kingdom, down from 13,000 in 1981. The reason for this and the research basis are discussed below.

Figure 9. Typical zebra crossing in the United Kingdom.

Figure 9. Typical zebra crossing in the United Kingdom.

"There is little difference in the average rate of personal injury accidents at Zebra and signal-controlled types. At individual sites, however, the type of crossing selected may have considerable effect on the future accident record." (DOT, 1995c).

Broadly speaking, Zebra crossings are considered inappropriate on high speed or high motor traffic flow roads, particularly multi-lane roads. The DETR guidance recommends that Zebras should not be installed on roads where 85 percentile speed is greater than 56.35 km/h (35 mi/h).

Since 1991 Zebra crossings can be raised, i.e. combined with a flat-top road hump, to produce a humped Zebra. This makes it easier for pedestrians with pushchairs, trolleys, wheelchairs, etc. to cross and helps to reduce motor vehicle speeds and emphasizes that drivers should give way to pedestrians crossing. However, the number of such crossings is still small.

Zebra crossings generally cause far less delay to pedestrians than Pelican crossings (Hunt, 1997). They are considerably cheaper to install and maintain than signal-controlled crossings. However, there has been a tendency for traffic engineers to replace Zebras with Pelicans and to choose Pelicans rather than Zebras when installing new pedestrian crossings. This is because of a number of factors. Firstly, as traffic flows have risen, there has been a considerable increase in the number of signal-controlled junctions and UTC systems. Signal-controlled pedestrian crossings (i.e., Pelicans) are seen as more appropriate in this environment, on the assumption that drivers are concentrating on signals to indicate stop rather than other visual cues. Secondly, where pedestrian demand is heavy, Pelican crossings allow motorized traffic to continue to flow. Thirdly, there is sometimes a perception, by both public and engineers, that Pelicans are safer and better than Zebras because drivers are controlled by signals rather than using their discretion.

When accidents have occurred at Zebra crossings, there has been a tendency to replace the Zebra with a Pelican in the hope that this will solve the problem. It certainly demonstrates that something has been done but it does not necessarily improve safety (or convenience) for pedestrians.

There are some signs that the Zebra may be making a come-back, particularly within traffic calmed areas, because Zebra crossings give pedestrians greater priority and are less visually intrusive and less expensive. Also, it is argued that Pelican crossings encourage the driver to look for signals and not for pedestrians and that this can have a detrimental effect on pedestrian safety.

Edinburgh has recently installed three Zebra crossings on arms of three busy roundabouts in the city center (George Street). Pedestrian flows are high. These are the first Zebra crossings to be installed in Edinburgh for 30 years. The reasons for installing Zebras were that they gave greater priority to pedestrians and reduced pedestrian delay; also that Pelican crossings would have been difficult to install in these locations. The performance of the Zebras, in terms of safety, delay, etc., will be monitored. Initial reports from officers indicate that they are working well and that, when motor vehicle queues build up, pedestrians stop to allow the vehicles to proceed, an interesting reversal of the usual priority.

York, another UK city known for promoting walking and pedestrian safety, has also begun to reverse its earlier policy of replacing Zebra crossings with Pelicans. This was being done to integrate pedestrian crossings with the UTC system, in order to reduce vehicle delay. Two humped Zebra crossings have recently been installed, partly because they give pedestrians greater priority and also because, being considerably cheaper than Pelican crossings, they can be installed more widely. Zebra crossings have also recently been installed in Norwich.

5.2 Pedestrian refuges

Since 1990 there has been a dramatic increase in the installation of refuge islands, particularly pedestrian refuges. (Local highway authorities already had powers under the Highways Acts to install refuges to assist pedestrians, but the 1991 Traffic Calming Regulations permitted them to also install refuges for non-pedestrian purposes.) This increase has been due to efforts to reduce pedestrian casualties and to assist pedestrians to cross roads on which traffic has grown substantially. Pedestrian refuges have also formed part of traffic calming schemes and local safety schemes designed to provide greater separation between oncoming vehicles and to reduce overtaking accidents. However, to a large extent, pedestrian refuges have been installed in preference to other traffic calming and road safety measures because they are easy to install. They do not require a Traffic Regulation Order, and there is no requirement to consult the public or others before installing. They require less signing and are generally less controversial or constrained by regulations than road humps. There has been a tendency, therefore, to install traffic islands and pedestrian refuges in response to perceived needs to do something.

Pedestrian refuges can provide a series of crossing points along a road where it would be impractical to install Zebras or Pelicans at each crossing location. The minimum recommended width (across the road) for a refuge is 1.2 m, but 2 m is preferred to accommodate wheelchairs, pushchairs, and cycles. Where pedestrian flows are high, this may need to be increased, although in such cases another form of crossing (Zebra or Pelican) may be more appropriate.

Detailed research into the effects of installing pedestrian refuges was undertaken by Thompson et al (1990) in Nottingham. A survey of 32 sites showed significant decreases (p<0.1) in 85 percentile speed at 9 sites and significant increases in speed at 4 sites (although all increases were at one scheme). Somewhat surprisingly no relationship was found between the residual width of the road (between 3 m and 4.5 m) and the proportion of vehicles exceeding the 48 km/h (30 mi/h) speed limit. Although the schemes were introduced to improve pedestrian safety, there was a slight increase in pedestrian accidents. This was presumably caused by an increase in pedestrian activity but this aspect was not measured nor were traffic flows. Regarding the change in accidents of all types, "a statistically significant reduction was only achieved at two of the schemes. In addition, the reduction in accidents at all of the schemes combined was not significant when compared to accident control data." Residents perceived that pedestrians' safety had improved (which is quite compatible with increased accident numbers if pedestrian activity increased) but perceived that safety for pedal cyclists and drivers was reduced.

Research into the effects of road narrowings, including pedestrian refuges, was carried out because of concerns about effects on pedal cyclist safety (Davies et al, 1997). This found that although there appeared to be a relationship between motor vehicle speed and carriageway width, other factors, such as traffic flows, pedal cycle flows, and congestion, were probably more significant than width in influencing speed.

Overall, it seems that pedestrian refuges assist pedestrians to cross roads more easily, with less delay and greater perceived safety. However, vehicle speeds are not necessarily reduced and pedestrian accidents may not reduce if pedestrian activity increases. There may also be adverse effects such as parking problems and problems for pedal cyclists. Unfortunately, as with so much road safety engineering research, the studies do not include exposure data so an overall assessment of the safety benefits is difficult to obtain. It may well be that pedestrian refuges do have greater safety benefits for pedestrians than the Nottingham accident data imply.

5.3 Curb build-outs

Curb build-outs may be used in isolation or in conjunction with other measures such as Zebra crossings or pedestrian refuges. They increase the prominence of pedestrians, particularly if there is curb-side parking, and reduce the crossing width. A study of an early scheme in Nottingham (Thompson and Heydon, 1991) where pedestrians were often masked by parked vehicles, found a reduction in average pedestrian accidents from 4.7 per year to 1 per year. The build-out extended 2.5 m into the carriageway and included substantial lengths of guard rail.

Most build-outs away from parked vehicles are considerably narrower, usually less than 1 m (3.28 ft) into the carriageway. As build-outs are often part of more comprehensive measures, specific evaluation of build-outs in isolation has been limited. As with pedestrian refuge islands, build-outs can cause concern to cyclists who are forced closer to motor vehicles.

5.4 Flat-top road hump

Many traffic calming schemes have been introduced in the United Kingdom in the past 5 years. These often include flat top road humps which make crossing more convenient for pedestrians and potentially safer by concentrating pedestrian crossing movements, particularly when parked cars may mask children crossing. They may be used in conjunction with curb build-outs. Flat-top road humps, like pedestrian refuge islands, give no precedence to pedestrians over vehicles. However, unlike the situation at refuges, drivers often give way to pedestrians as if they were approaching a Zebra crossing. This is because vehicles need to slow down for the road hump. It is more likely to happen where traffic is queuing and the number of pedestrians crossing is high. These schemes have been criticized for introducing ambiguity: some pedestrians behave as if they have some legal priority and while some drivers give way, by no means all do so. However, they are generally successful in that they provide pedestrians with safer crossing locations that are easier to use (particularly for those pedestrians with push-chairs (buggies), shopping trolleys, or wheelchairs) and reduce pedestrian delay.

Since the Road Humps Regulations 1996, local highway authorities in England have been permitted to install road humps without a prior speed reducing feature, although DETR generally advise against doing so.


Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101