Honest John Real MPG is back: real-world fuel economy for 776 models, free

Real MPG was the feature Honest John readers used most and talked about least, right up until August 2026 when it vanished along with the rest of the site. It's back, rebuilt from the archived averages, the official NEDC and WLTP figures for every engine, and new submissions from readers, so you can look up what a car really does to the gallon again.

Is Honest John Real MPG back?

The data is, on our free Real MPG section. Honest John itself closed in August 2026 and every honestjohn.co.uk address now redirects to Drive.co.uk, which didn't take Real MPG with it. So we rebuilt the section from three sources: the averages the archived Real MPG pages held, the official NEDC and WLTP figures for every engine, and new submissions from readers, which are what keep the averages moving. It's free, it needs no account, and it stays that way permanently, because running-cost data only works if everyone can check it. We're not affiliated with Honest John or Volkswagen Financial Services, and every page credits the archived source it came from.

One thing didn't survive and never will. The individual driver submissions behind each average were never published anywhere, so nobody can resurrect the submission history. What we saved is the output: the per-engine averages the site displayed, which is the part you actually read.

What's in the rebuilt Real MPG database?

776 models across 48 makes, 510 of them with a page of their own, built on more than 319,000 owner submissions. Each model page splits by generation, and each generation lists every engine we hold a figure for, with four things side by side:

Models with thin data sit on their make page rather than getting dropped, so a car with one engine and a handful of submissions is still findable. Every page links to the archived Honest John capture its figures came from, because credit matters and because you should be able to check us.

What does the "% of official" column actually tell you?

It tells you how honest a car's brochure figure was, and the spread between engines is wider than most people expect. Two real examples from the database, both with thousands of submissions behind the generation:

• The Ford Fiesta 1.0T EcoBoost 100, from the 2013 to 2017 car, has an official figure of 65.7 mpg. Owners average 46.3. That's 70% of official, and the gap is roughly 19 mpg.
• The Volkswagen Golf 1.6 TDI 115, from the 2012 to 2020 car, averages 54.4 mpg against a WLTP figure of 52.3 to 57.8. That's 99% of official.

Both cars are doing exactly what you'd expect them to. The Fiesta's little turbo petrol was tested on a gentle laboratory cycle it flatters, and it needs a light right foot to get near that number in traffic. The diesel Golf spends its life at the sort of steady speeds its test was measured at. Neither figure is a scandal, and neither car is faulty. It's just that "65.7 mpg" and "52.3 mpg" on two brochures don't mean what they look like they mean when you park them next to each other.

That's the whole argument for a real-world column existing. Official figures compare cars fairly against each other under identical conditions. They're a poor predictor of your fuel bill, and the size of the error changes by engine. Rank the database on what owners really get and you end up with a very different list of the most economical used cars from the one the brochures would give you.

Why do some Real MPG figures have an asterisk?

Because they're measured against the WLTP official figure rather than the older NEDC one, and the two aren't comparable. NEDC was the test cycle until it was phased out from 2017, and it was generous. WLTP replaced it and sits much closer to real driving, so a WLTP official figure is a lower number for the same car.

The Golf above shows why this matters. Its 54.4 mpg average is 99% of the WLTP figure and about 80% of the NEDC one. Same car, same owners, same road, two very different-looking percentages depending on which test you hold it against. So when you compare percentages across our tables, compare like with like: an asterisked figure at 95% and a non-asterisked figure at 95% are not telling you the same thing about the car. There's a lot more in why real MPG comes in under the official figure, including what the whole database says a normal shortfall looks like on each test.

How do I look up a car's real-world MPG?

Go to Real MPG, pick the make, pick the model, then find the generation that covers the year you're looking at. The engine tables sit underneath. It's quicker than the route we used to have to recommend, which was hunting for an archived copy of the old page, and it's the same underlying data with the official figures added.

If you'd rather see the original, the Internet Archive still holds most of the old Real MPG pages. Search web.archive.org for the old address, which followed the pattern honestjohn.co.uk/realmpg/ then the make then the model, and pick a snapshot dated before August 2026. Images often don't load in archived copies. The numbers do. We link the exact capture we used at the foot of every model page, so you can go straight to it rather than hunting.

For the official half on its own, the VCA car fuel data tool is free, official and covers every car sold new in the UK. It won't tell you what anybody actually got.

What if the real figure is miles below the official one?

A big gap usually means the test flattered the engine, which is normal and is what the whole table is for. But an unusually big gap on one specific engine, when the rest of the range looks sane, is worth a second look, because a few known faults show up as fuel economy before they show up as anything else. A sticking DPF regeneration cycle, a tired injector, a failing coolant thermostat that never lets the engine get properly warm: all of those cost you fuel first and money later.

That's the other half of what we rebuilt. The known-fault data from Honest John's Good & Bad pages now lives on our common problems pages, so if a car's real-world figure looks off you can check whether the model has a documented reason. The Ford Fiesta known faults and VW Golf known faults pages are a good place to see the shape of it. Both sections sit together on the free car data hub.

How do I add my own car's figure?

Every model page has a short form at the bottom. Pick the generation, pick the engine, put in what your car genuinely returns, say whether you're mostly town, mixed or mostly motorway, and add your annual mileage if you fancy. There's no postcode, no email and no account. Submissions get checked and folded into the averages in batches, which is exactly how the original got built over 25 years.

Two things worth saying about that. Give the figure your car really does, brim to brim or off the trip computer over a decent stretch, not the best tank you ever had. And put in the boring cars. The database is thickest on Golfs and Fiestas and thinnest on the models nobody thinks to submit, which are usually the ones somebody is about to buy blind.

Can a check tell me what a specific car will cost to run?

Partly, and it's worth being straight about which part. On the fuel itself, no check beats the crowd data above, ours included, because economy depends on how and where you drive more than on anything in a database. And on the provenance side we get the same records as everybody else: finance from the lenders, theft markers from the police, write-offs from the insurance industry. Nobody holds a private source.

What we do that a plate lookup doesn't is read the advert. Every other check starts and ends with a registration, so the seller's own wording is the one part of a used-car listing that nobody examines. Our free scan takes the advert text and reads it against the car's full DVSA MOT history and the known faults for that exact model and engine, then hands back a verdict, a risk score and the single biggest thing it found. No card, no password. So if the advert says "just serviced, drives faultless" while the MOT history has been carrying the same advisory for three years, that contradiction is the finding, and it's a contradiction only something that has read both halves can spot.

Here's what it can't do, and no product can. It can't tell you what you'll get to the gallon. It can't see unrecorded accident damage or the quality of a repair, because those aren't in any database at any price. And it can't drive the car for you.

So: use the Real MPG figures to work out whether the model suits your mileage, then run the free scan on the actual advert before you drive out to see it. If it turns up something worth digging into, the free report upgrades on the same link later without re-running anything, so starting free costs you nothing even when the answer is walk away.

The short version

Real MPG isn't gone. It's rebuilt, free, at /real-mpg: 776 models, 48 makes, more than 319,000 owner submissions, with the official NEDC and WLTP figures alongside so you can see the gap for yourself. The original driver submissions are gone permanently and nobody can bring those back. Official figures are still fine for comparing cars and still optimistic for predicting your spend. And the thing that made Real MPG work in the first place, drivers putting their own numbers in, is open again.