I’ll start this with the graph itself, and then go into a few notes below:



Cycling – people won’t do it because of hills, right? Or more generally, the Dutch all cycle because “Holland” is flat?

Yes, I know, there is no Holland, it’s the Netherlands, and this matters, because the hillier bits are well outside the two provinces of North and South Holland which some people seem to think makes up the whole country. As with many other myths, “Holland is flat” also starts with an element of truth. It’s not entirely fair just to say that hills aren’t an issue, or that the lack of hills in many parts of the Netherlands isn’t part of the reason why so many people cycle there. But ultimately, that’s all it is – part of the reason, and no more.

Let’s look at this from a few logical standpoints.

  1. If flatness was the only factor which determined cycling rates, then other places which were flat would also have very high levels of cycling. We only need to find one such place to disprove this, and as it happens, somewhere with a similar population to the Netherlands, which is almost entirely flat, and which has two globally known cities will do just fine. That place is the US state of Florida. Oh of course, Florida is hot (as the Netherlands is windy), but the real reason for the general lack of cycling in Florida is almost entirely down to the lack of provision.
  2. If hills really put people off from cycling, then we only need to find one city with hills that has a reasonable level of cycling, and then we can learn from there.

Flatness – a factor and no more

There is no doubt that flat terrain means that less mechanical effort is needed to move a machine like a bicycle across the same distance. However, it’s equally true that terrain is just one of numerous different factors which determine which choice of transport any individual or group will use for any one journey. If hills are going to be a problem in Coventry, then those same hills in Germany or Belgium (yes, Belgium isn’t flat either) are going to be the same problem, assuming relatively similar climates.

Yet the real reason why Dutch cycling rates are so high is not because of the lack of physical effort needed to move around Dutch cities, but because of relative lack of mental stress needed to move through the street environment. Physical exertion tends to mean that people may try a different approach (and for increasing numbers, this means using e-bikes), whereas cycling in unpleasant traffic tends to mean people just won’t cycle at all.

This is where the Netherlands kept a relative advantage in the 1970s. Just as Dutch cities were tearing themselves apart to cater for the car, our British cities were doing exactly the same thing. The saving grace for the Dutch cities was that although the cycling mode share fell, it never dipped below 10% in most cities, thus there was much more of a grass roots movement to spearhead the protests against the road deaths which then followed.

So to say this once more – yes, the flatness in the Netherlands could be attributed to meaning that more people tolerated otherwise unpleasant road conditions in the 70s, but it did no more than that. Almost all of the growth in cycling since then can be attributed to changes in the infrastructure (the liability laws didn’t come until the mid-90s, long after mass cycling had been restored, not before it).


Florida is flat, but nobody cycles here. If you then move on to say it’s hot, then you are just making up another excuse.

Hilly Cities With High Cycling Levels

If we can find one city with reasonably high cycling rates, then we have something to show that hills really aren’t a problem. This doesn’t have to be a British city, because all we are demonstrating is that hills are not a barrier to cycling. In order to do this, I combined two sets of information into a single table:

  • Which cities in a global list of top cycling cities are also known to have hills?
  • What is the cycling rate in each city?
  • For each of these cities, determine a score to show how hilly they are.

Note that I am only aiming to look at cities which are (a) reasonably hilly and (b) which have relatively high cycling rates. For this, I’ve set the bar at at least double what Coventry has now, or 6% of trips. I’ve also included Cambridge in the comparison, firstly as the leading British cycling city, but also because it is not quite as “pancake” flat as certain Dutch cities. I also looked at mode share in a number of other cities, but excluded them as their gradient index was zero. These are – Copenhagen, Oulu (Finland) and the Dutch cities of Amsterdam, Rotterdam, Groningen and Utrecht.

I’ve made no attempt at all to assess the standards of cycling infrastructure in any of these cities. Doubtless to say, having been to both Cambridge and Maastricht, I know that these “leaders” in this table still have plenty of room for improvement. Arnhem may be a reasonable indicator of the relationship between hills and cycling rates, because it most certainly has gradients, and is hillier than, for example, Edinburgh, using the criteria given. However, even with Arnhem, there has been a significant investment in recent years, especially with the new fast link to Nijmigen, suggesting that the provided mode share of 19% may well be very conservative.


Cycling up Tongerseweg in Maastricht. This photo was taken on a quiet Sunday afternoon, yet there are still more people moving on bikes than in cars.

So do hills really impact cycling rates?

This is a selection of cycling cities with hills, not a random selection of cities, but it is taken from a list of the world’s leading cycling cities. Given this list, there is still a trend which shows that higher cycling rates exist in cities which are flatter, yet I could just have easily found a load of flat cities with no cycling to balance the whole thing out, but this was not the point.

This chart is not presented as a statistical comparison, but simply as a way to show that hills are not a barrier to the development of mass cycling. If positively alpine Innsbruck can have 4x the cycling rate of Coventry, or if slightly more rolling Maastricht can have cycling rates which are 10x that of Coventry, then clearly hills are no impediment to the development of everyday cycling.

Therefore, if “Holland”” is flat, or if Cambridge is “flat”, then so is Coventry!


INNSBRUCK            100 14
BRISTOL               55 9
BERN               52 16
STAVANGER               31 6
ARNHEM               29 19
DRESDEN               25 15
EDINBURGH               23 10
MAASTRICHT               23 27
COVENTRY               14 3
CAMBRIDGE             4 32


Table Notes

  • Cities selected from a list of top 100 cycling cities known to have hilly terrain. Additional inclusions include Bristol (hilly with rapid cycling growth), Stavanger (hilly with relatively recent major cycling investment), and Cambridge as the UK’s leading cycling city (generally, but not quite entirely flat).
  • Cycling mode share figures taken from City Clock article (2014), unless more recent figures available – http://www.cityclock.org/urban-cycling-mode-share/#.WD9PFLKLQuU

Determining the Hills Index

  • For each city, google maps cycling route planner was used to select four routes, roughly running north, east, south and west from the main railway station in the city.
  • The railway station was used as this provided a central point, rather than trying to otherwise define one. In the case of Edinburgh, routes were determined from Princes Street, and do not include the steep climb out of Waverley Station.
  • Each route selected was roughly 5km or 3 miles, a distance that is reasonable to cycle unaided on level terrain by a reasonably fit person, wearing everyday clothes.
  • Gradient was determined by dividing the height gain by the total distance of the route. Downhill running was not included.
  • Generally, journeys away from the main railway station would be uphill. Where they were downhill, the reverse direction was considered.
  • Routes determined by google maps to be “mostly flat” were given a gradient of zero.
  • It is noted that “mostly flat” does not mean no climbs at all. For example, Coventry City Centre to Inverness Close is marked as “mostly flat”, but it includes a steep rise of 20m at the end. Irrespective of this, averaged out over 4 routes, Coventry is not considered to be a hilly city.


Leave a Reply

Your email address will not be published. Required fields are marked *