bollards or chicanes on cycle routes are dangerous

Using bollards or other obstacles as "speed calming" on cycle routes will create more injuries than it averts.

In some four hundred kilometres of cycling around the Netherlands in May 2026, within cities and in between them, it was remarkable how few bollards - or other obstructions - there were on any cycle routes. They are basically used only where necessary to block motor vehicle access, and then take the form of the minimum number of hi-viz bollards, of such a height that they don't risk striking handlebars.

I can't show you photos of things I didn't see, but this set of chicanes in Scheveningen was the most egregious exception.

The cycle tracks and lanes in the background are Dutch, but this chicane seemed very British, especially with the "this isn't going to work anyway" desire line around it

In the UK, bollards and chicanes are often used to try to keep out motorcycles. The argument against this is clear: doing that will also block access for "many cycles, mobility scooters, wheelchairs and disabled pedestrians" (Wheels For Wellbeing).

Here I'm concerned with a different problem, which is the use of bollards as a safety measure to slow people cycling down. I look at three recent examples from Oxford (all put in in the last year).

Donnington Recreation Ground

At the western end of a new path across Donnington Recreation Ground - just put in just, in early 2026 - we find these three bollards.

the western end of the path across Donnington Recreation Ground

These are not necessary to keep out motor traffic - there are bollards at the other end of the new path, and also on both ends of the section of Meadow Lane behind these bollards - but were put in to slow people cycling. ("The entrance would feature bollards to reduce the speed of cyclists", says the planning approval). But it is hard to see these bollards doing anything at all to slow people cycling, and they are also a hazard. Maybe once in every million trips someone - at night, or trying to cycle side-by-side with someone, or drunk, or just unwary - is going to hit one of these.

Moreover, by reducing the space people walking and wheeling and cycling have to manoeuvre in, these bollards actually increase the risk of both cycle-pedestrian and cycle-cycle collisions (though on a relatively low-traffic route like this, multi-user interactions will happen much less frequently here than simple "cycling past bollard" events).

Flower Lane

Flower Lane, put in in late 2025

This is a huge improvement on the previous layout, where a fence blocked direct movement from Flower Lane to Howard St. But... In response to concerns about pedestrian safety, an attempt has been made to divide cycling from walking, but the width available is too little for that. Observation suggests the painted guidance is largely ignored — and it is actually safer to have people cycling keep to the left, because that's the expected behaviour everywhere else.

And again we have some unnecessary bollards. The old wooden bollard at the back suffices to stop small cars trying to get through, and the two bollards on the right stop cars parking. But the other two are gratuitous, and actually make the space less safe by reducing the space people walking, wheeling and cycling have to move around one another. (There was originally one more bollard, just behind the front one.)

Wren Rd

these look nice, but are a safeguarding failure as they will result in completely unnecessary injuries to children (and others)

At the end of Wren Rd, outside Cutteslowe Community School, a set of pencil bollards were put in at the beginning of 2026, not to restrict cars but to slow people cycling ("the positioning of the bollards at the junction of NCN 51 and Wren Road were also adjusted to help reduce cycle speeds outside the school entrance and make it safer for pupils entering and leaving the school at the start and end of the school day", "Wren Road - proposed parking amendments formalisation" consultation)

First of all, as with the other examples, this is really not going to slow people cycling down, or if it does it will slow down already slow and less confident people, not anyone comfortable cycling at speed. Secondly, it's a clear hazard - the risk of someone hitting one of these bollards and injuring themselves is non trivial, so over time, on a reasonably busy cycle route, that is going to happen. And thirdly, why is it necessary to slow people cycling here at all? The stub of Wren Rd hardly has speeding motor traffic on it and visibility is excellent, so if it does get crowded at school-run times, people cycling can and will just slow down anyway.

This is, of course, not a new problem. Just behind the last photo is the Cutteslowe bridge, which has two sets of chicane bollards which are both hazardous and unnecessary, if not pointless. These were presumably put in with the bridge in 2000.

Is this a real problem?

But is this a real problem, I can hear people asking? I haven't seen any reports of injuries from these three locations yet, but the bollards there haven't been in place long. In September 2025 I saw this report of an injury from hitting the bollards on Barracks Lane:

"A friend clipped one of these brackets that are either side of the bollard in the dark, fell off her bike, and broke her elbow. The bollard itself is marked with reflective stickers, but the brackets aren't, and are invisible at night."

Which elicited this comment:

"This is far from the first time. I passed there some time last year(?) and there was a group of student aged folks waiting for an ambulance because one of their number had done exactly this. Suspected shoulder area fracture, iirc."

Those are the reports that got me thinking about this, but while writing this (in June 2026) I sent a query to some local groups — perhaps a hundred people, but certainly a small fraction of the number of people cycling in Oxford, even if expanded to include family members — and got these responses:

"I cycled into a bollard on the South Parks Road cycle path, went over the handlebars, but wasn’t too badly hurt thankfully. The bollard was in the middle of the lane by one of the departmental exits, presumably to obstruct wider vehicles from using it."

"Marston Ferry Rd, leaving Swan School towards Cherwell Drive. My daughter caught her handle bars in the too near sticky out bollards and gashed her leg and elbow. Put her off biking for a week but nothing serious."

"My daughter broke her hand when she cycled into a bollard on the MFR cycle track coming home from work one (dark) night - that was 8 or so years ago."

"[My son] once fell of his bike after clipping the bollard at the South Parks Road end of the Marston Cycle Path. Didn't spot the handles on the sides in the dark. Injuries: scraped palms and knee."

"I got heavily bruised when I came off my bike back in January thanks to the chicane in place on Meadow Lane where it crosses Bedford Street by the entrance to Mary and John school. To avoid ending up off the path in the mud there when heading south, you have to perform a sharp turn left and then right after going round the barrier. The asphalt surface is smooth so, as I found, a light frost can be enough for your bike to slide from under you as you turn."

"I saw a cyclist crash into the bollard/lamppost outside the Hayfield Rd deli."

There are a lot more injuries involving fixed obstacles than are ever reported - around 90% of all cycling injuries, especially those resulting from single-cycle collisions, are never reported to either the police or health services. (So injury-prevention measures disproportionately focus on multi-user collisions.)

On a well-used cycle route, anything that creates even a one in a million chance of a serious injury will generate one every few years. (In most cases - and certainly in the three I discuss above - the bollards are also hazardous to pedestrians, though collisions there are going to be lower speed and likely to be less severe.)

Conclusions

• Putting bollards or chicanes or other obstacles on cycle routes will always create danger and increase injury risk.
• Obstacles are unlikely to slow people cycling down much if at all, unless they also reduce accessibility (and impede some people's use of the route).
• In most cases, slowing people cycling down is unnecessary and serves no clear purpose. In none of the three examples above, for example, does it seem useful.

Recommendations

• We should avoid using bollards (or chicanes) to try to slow people cycling, as they will almost always generate more injuries than they avert.
• Where bollards are necessary to block motor vehicles, the minimum number necessary should be used. It may not be necessary to have bollards at both ends of a stretch of cycle track, for example.
• Where we have bollards on or adjacent to cycle routes, they should be designed to minimise hazards. So they should be hi-viz, be 0.8 metres rather than 1 metre high to reduce the risk of hand or handlebar strike, and so forth.
• Where there are potential conflicts, alternative safety measures should be prioritised. These might include improving visibility, redirecting and simplifying flows, increasing the space available, and slowing down motor vehicles if they are involved.

bollards in Groningen, preventing motor vehicle access while creating the least danger and stress for people cycling