← survey home · UK climbs · US climbs · Swiss climbs · Austrian climbs · French climbs
What about the road near me?
The most common question this survey gets is some version of “there’s a brutal hill near me — why isn’t it here?” Usually the answer is not that we missed it. It is that the road is genuinely hard and genuinely not steep, and those are different things.
Hard is not the same as steep
A gradient is a ratio: metres up per hundred metres along. It says nothing about how much the road hurts. Length, wind, surface, traffic, your gearing, how far into the ride you are and what came before all change the effort without moving the number by a fraction of a point. A long drag at 3% into a headwind is real work. It is still a 3% drag.
The reverse is also true. A 20% ramp that lasts eighty metres barely registers as effort if you hit it with speed, and it will still be one of the steepest things in a county.
A worked example
Take a road that feels like it has hills in it: US Route 1 through Topsfield, Massachusetts. Measured the same way as everything on this site — 1 m LiDAR, sampled every metre along the carriageway — 7.6 km of it looks like this:
| Measured over | Steepest, northbound | Steepest, southbound |
|---|---|---|
| 100 m | 7.2% | 7.2% |
| 250 m | 7.0% | 7.1% |
| 500 m | 6.2% | 6.6% |
The longest continuous rise is 1.8 km for 53 m of ascent — an average of 3.0%, with one 7.2% hundred inside it. End to end the road climbs about 138 m in one direction and 143 m in the other, in a series of rollers, and none of it is steep by this survey’s standards. It has no page here. Ride it into a north-easterly and you will still feel like you climbed something, because you did: 143 m of ascent is 143 m of ascent whether it arrives at 3% or at 20%.
That measurement was a one-off, done to answer exactly this question, and it is not part of the published dataset.
Your eyes are not instruments — and neither is Street View
Slope perception is famously, consistently wrong. People asked to judge a hill by eye overestimate it by a factor of three or four: a 5° slope — about 9% — is routinely called 20°. The error is not carelessness. It is how the visual system works, and knowing about it does not switch it off.
Photographs inherit the problem and add their own. A camera pointed up a climb with a long lens compresses the road into a wall. The same road shot wide, from the saddle, flattens into nothing. Street View is especially misleading in both directions: the panoramas are levelled when they are stitched, so the frame keeps the horizon horizontal and the road’s slope relative to that frame largely disappears. A 20% wall can look like a mild rise on screen. Then, on the next road, a camber or a sloping verge tilts everything and a 4% drag looks vicious.
The strongest illusions come from the surroundings rather than the road. If the land either side slopes more steeply than the tarmac, the road reads as flat or even downhill against it — that is the whole mechanism behind every “magnetic hill” where a car appears to roll uphill. A road can look like a climb because the valley wall behind it is one.
This is the entire reason the survey exists. Not because people are bad at judging hills, but because there was no falsifiable number to check them against.
So what counts as steep here
Every figure on this site is a window maximum: the steepest stretch of exactly that distance anywhere on the climb, stated with the distance. That is the only honest way to compare roads, and it is why a road cannot be dismissed or promoted on a single number — see how to read these numbers.
The first sieve is deliberately generous: a road is scanned onward only if it reaches about 7% over a 200 m window and climbs at least 20 m. Most roads that feel steep fail the first test comfortably. Roads that pass then have to survive the checks that decide whether there is a climb there at all — where it starts, where it tops out, whether the steep part is the road or a bridge deck, a tree canopy or a gap in the data.
Telling us about a road
The report form takes two quite different things, and they get different answers.
A road you think we have missed. These we collect. Suggestions go onto a list that feeds future survey work, and the list is genuinely read — but it is a list, not a queue, and we do not promise that any particular road will be measured, or say when. Most suggestions turn out to be the honest case this page is about: a hard road that is not a steep one.
Something wrong on a page that already exists. That is a different matter, and worth telling us: a line drawn down the wrong side of a carriageway, an extent that starts or stops in the wrong place, a figure you think a bridge or an overpass has inflated. We do get these wrong, the fix usually applies to more pages than the one you noticed, and we work through them as we can.