How Autocar tests cars

It’s common for automotive journalism brands to crow about how thoroughly they assess new cars, but only one can claim to have been doing it for almost a century.

Autocar delivers several kinds of reviews on new cars, with different levels of technical and functional scrutiny. 

Our shorter reviews typically address cars that we’re driving for the first time - either in final production form or as late prototypes. Time in the car can be limited here, technical and equipment specification might not be finalised and some systems software may not be to production specification. For those reasons, these reports don’t have star ratings. 

Once we have driven a car in the UK, in right-hand-drive form and typically at greater length, we will rate it, either as part of a shorter UK test report or as the subject of a full, instrumented Road Test. 

In the latter case, we take empirical performance and refinement benchmarks from it at Horiba MIRA Proving Ground, as well as subjective limit handling impressions, plus key measurements of practicality, efficiency and kerb weight. All UK tests and Road Tests receive proper verdicts with star ratings informed by everything we have learned and the combined impressions and experience of several testers.

The Autocar Road Test first appeared in print, so titled, in April 1928; but Autocar’s history of publishing review articles with accompanying empirical performance, handling and drivability tests is older still. And today, while our testing methods have developed and changed along with the cars they’re intended to scrutinise, the totally rigorous, fiercely independent and studiously objective approach they represent remains unequalled in the field.

For electric cars, we conduct a DC rapid charging performance test, as part of which we produce a weighted average rapid charging test result that can be compared with the manufacturer’s claimed peak charge rate. That gives a more useful indication of how fast any EV can actually charge across the full breadth of its battery condition; and how long you might therefore be waiting for a fully restored range.

All of the data we generate is presented both in print and online, and it feeds into the most committed, repeatable and fair effort we can make to understand and evaluate any car before we pass judgement upon it. Autocar’s team of experienced road testers then compare impressions and agree on a verdict that represents their combined voices, and that of the Autocar brand as a whole, taking into account the abilities of key rival models.

As a result, no road test is more authoritative and robust, meaning when Autocar delivers its test verdict, it’s one that you can trust.

Our marquee comparison tests, meanwhile, such as Britain’s Best Driver’s Car, involve in-depth back-to-back testing on both road and track; often taking several days to conduct if the field of cars is large; and usually result in a collegiate verdict arrived at by way of a judges’ scorecard.

Nobody takes the business of fitness-for-purpose-based new car assessment more seriously than we do; nobody goes to greater lengths to deliver a thorough, relevant, fair and definitive verdict; and nowhere have they been doing it for longer. Whether you want to know every detail about the performance of your next new car, or just to disappear into reverie about one you might never own or even drive, Autocar has you covered.

Read more: Why you can trust Autocar

How Autocar tests cars: Design and styling

Autocar can’t tell you whether you’ll like the way a car looks, any more than any other website can. That’s why we generally don't stray into the bounds of the subjective on car styling – however much you might trust our expertise to do so on your behalf.

If a clear consensus or shared reaction to a new car’s design does emerge on any test subject – or if there’s a particular mood about one held abroad in the car-facing public at large – it’s important to acknowledge it, and how it might affect your perception of a new car, or its all-round appeal. 

There are important design conventions for particular types of car; and it may be appropriate for us to comment on how well they’ve been observed – or, in some cases, reinterpreted or reinvented. Generally, however, the score you’ll find in this section doesn’t simply reflect how good-looking we think a new car is, because your opinion alone on that is the only one that matters. We will, however, always award credit for good objective design; for well-conceived and well-executed engineering; for bold technical innovation; and for demonstrably superior technical systems and solutions. 

Every car we test for the Road Test is weighed in running order, so that we can verify manufacturer kerbweight claims, and also get an empirical measurement of weight distribution – both key factors in a car’s performance, handling and how much you’ll need to pay in fuel, be that petrol, diesel, electricity or a combination of those. 

Fitness for intended purpose is kept in close focus at all times: so a new type of chassis that saves weight or adds rigidity in a sports car might win it praise, as might a more efficient type of electric motor in a long-range executive saloon, or a cheaper drive battery in an affordable EV. Moreover, heavy cars aren’t necessarily poorer ones: it depends on the brief they are to serve.

Our focus is always to assess whether the design, specification and mechanical configuration of the car in question sets it on a path to deliver against its own particular positioning and brief, and within the context of the market in which it sits. But we also recognise and reward what may make it special and different. 

How Autocar tests cars: Interior

Autocar believes that cars are, first and foremost, devices for driving; not just means of transport. That’s why, when we assess the interior of any new car, we start by taking a view on how well it accommodates the driver, how well-placed are its major controls and instruments, how much space it affords, and how clear is the visibility it provides in all directions.

Our common expectations of one car’s standard of passenger comfort may be quite different from the next, of course. It’s quite reasonable for a lightweight sports car to cater primarily to a shorter, slighter-built driver than a luxury SUV might, when everything else about that car is about particular leanness and weight-saving. But if the level of accommodation of any new car – either for one occupant, two, three, four, five, six or seven, or more – falls below the standards of its rivals, we will report it. And not in purely subjective terms.

Every new car in one of our multi-section reviews is measured in detail: for maximum and minimum front-row legroom and headroom, for typical rear legroom (which is measured behind the driver’s seat when the latter is itself set for 1000mm of legroom) and rear headroom, and for maximum and minimum cargo loading length, width and depth in the boot.

We also measure pedal and steering wheel placement relative to the centreline of the driver’s seat, so as to objectively recognise good and bad right-hand-drive primary control ergonomics – and the comfort and control levels that result. And, before we’re done with the tape measure, we use it to independently record a car’s overall width across the door mirrors; the span of its front doors when fully open (a useful indicator of how easy it is to access in a garage or tight parking spot); and the maximum height of its bootlid or scissor doors when fully open.

By the collection of all this empirical data, and because we compare it with that of so many rivals, you can trust us absolutely when we write that a new car is more or less spacious, practical or accessible than its competitors. And, because we also collate and record our subjective impressions of a cabin’s seats and wider comfort levels, its apparent integrity and solidity of feel, the richness and perceived quality of its fixtures and fittings, and the sophistication, clarity and usability of its digital infotainment and multimedia technology, our all-round assessment of a car’s interior puts you in the driver’s seat better than any other.

How Autocar tests cars: Engine & performance

In an era when electric cars are cutting acceleration times by significant margins, when manual gearboxes are becoming vanishingly rare, and automatics with launch control systems have taken the onus off car manufacturers to ensure their sports cars are easy to drive as well as fast, some say that the independent benchmarking of performance is becoming irrelevant. 

The reality is that the verification of vehicle performance is at least as important – and interesting – today as it ever has been; and Autocar approaches it with unrivalled attention to  detail, delivering you the dynamic testing benchmarks you can trust.

All of our Road Test subjects are fitted with Racelogic VBOX Sport satellite timing gear, which is used to benchmark standing start acceleration, in-gear acceleration (where applicable) and stopping distances from 30-, 50- and 70mph. We will use whatever electronic driver aids are fitted to the car, be that launch control, anti-lock braking, electronic stability control and emergency brake assist. But, since our tests are done the whole year round, we can’t guarantee optimal test conditions in every case – which are reflective of real-life driving conditions. Yet, whatever the test conditions were on the day will be recorded along with the results in order that reasonable allowances might be made by the reader.

Cars are figured with only the driver onboard, with a full tank of fuel, and with their engines, drivetrains, tyres and brakes at optimal operating temperature and pressure set to the manufacturer’s specifications. In the case of EVs, their drive batteries will be at optimal temperature and voltage also, with no test subject starting performance testing with less than 80 per cent state of charge.

Our brake testing takes some account of the effect of fade; the middle-sitting result of three concurrent maximum-pedal-pressure stops from 70mph is recorded, with braking starting from an indicated 80mph so that transient vehicle pitch doesn’t influence the results too much. But it isn’t intended to give a specialised assessment of how vulnerable a car’s braking system may be to overheating or fade-related deterioration. Our results deliver a verdict of braking performance you can rely on whether you’re on track, a motorway or driving to the shops.

Our in-gear acceleration testing provides some empirical indications of engine tractability and general drivability. But the impressions our highly experienced expert testers deliver following extensive road testing are far more useful in describing instances of turbo lag and other engine response irregularities, as well as how widely power and torque are spread; and, in EVs, how cleverly brake energy regeneration is handled. These factors are as important in a city car on the school run as a supercar on a race track or an SUV on a farm track.

In luxury cars and performance cars, subjective impressions of the richness of audible appeal and the inherent character and drama of power delivery are also noted. As time passes and powertrain technology changes, so the relevant standards such things can be judged must also change: but, clearly, no supercar ought to be without a sense of multi-sensory excitement to accompany its outright pace. 

How Autocar tests cars: Ride & handling

Several hundred miles of road testing informs the ride and handling assessment of every new car in an Autocar Road Test. Most of it is done on a mix of UK road surfaces; but some of it is done on dedicated handling tracks where the outright stability and controllability, limit handling characteristics and electronic traction and stability aids can be fully assessed. You can have confidence that any underlying dynamic traits, or electronically masked characteristics, will have been exposed to the light of day as a routine part of what we do.

The most performance-oriented new cars also set benchmark laptimes at a dry handling track at the hands of our highly-experienced testers, so we can take a view on how much effort, skill and commitment is necessary to drive them to their ultimate limit; and also, on how much lap speed they can carry relative to their rivals.

The focus of our attention is always the car’s intended environment, its role and what owners are most likely to ask of it. With a sports car, we may spend longer describing on-track handling manners; whereas with a mid-sized SUV our circuit element may be more of a formality. In both cases, though, a subject’s on-road ride and handling will always remain at the heart of our concern. No road car is as good as it should be if it can only be enjoyed on a trackday, after all; or if it only rewards its driver at speeds and effort levels that just can’t be approached in normal driving. 

But with the majority of the cars we test, we will be overwhelmingly concerned with the fundamentals of good vehicle dynamics here. Whether a car steers and behaves consistently and as you expect it to; how agile and manoeuvrable it is, and whether its size is a factor; whether it simply ‘goes where you point it’; and whether it feels accurate, stable and contained on its suspension at higher speeds, or imprecise and out-of-its-depth. 

Part of the ride and handling section is also devoted to the comfort, refinement and stability of the car we’re testing. That’s assessed both objectively – via in-cabin decibel readings taken at idle, 30mph, 50mph, 70mph and at maximum revs in either 3rd or 4th gear – and subjectively by the road test jury’s impressions of the car’s various sources of noise, and how well those sources are filtered.

How Autocar tests cars: MPG and running costs

This section is more nuanced than you might think. It takes a view on all of the things that define and inform the value and affordability of a new car, as well as those affecting cost and ease of ownership. And, usually, better value is better. 

But it also has to reflect the fact that, in some echelons of the car market, a keen price might conceivably be a turn-off to buyers. For the likes of Rolls-Royce, Ferrari, Bugatti and others, value is defined differently; and exclusivity and customisation matter so much more.

The showroom price of a particular car can be a misleading place from which to judge it; which is why we take a view also on residual values through forecasting from CAP HPI to see what it will be worth when you come to sell, and of both business contract hire rates and manufacturer personal finance rates, before deciding where any new car sits on value for money. 

With fleet-run EVs and PHEVs, benefit-in-kind tax qualification – which in some cases is defined by electric range – is important, too, and will inform our thinking. For cars more likely to be bought and run privately, meanwhile, fuel efficiency may still be king. You can trust the Autocar road test desk to have the expertise to know the difference.

And so, both EVs and ICE cars, and performance cars and family hatchback alike, we end up with a reliable indicator of what economy the subject will return on a longer touring trip; what it will return in mixed day-to-day use (admittedly with some intensive driving included); and what it might average on a trackday or circuit test. The results are all indicated by the subject’s trip computer, for reasons of practicality; but our methodology is the same for EVs as it is for piston-engined vehicles.

We also have a dedicated test of DC rapid charging performance for electric vehicles, as part of which every subject is plugged into a rapid charger capable of supplying enough power to meet the car’s claimed peak flow rate, and we record how much electric power it actually draws as it passed 10-, 30-, 50, 70- and 90% state of charge. A weighted average charge rate is then calculated, giving a reliable indication of how long any EV might make you wait for it to take on power during a longer trip. 

How Autocar tests cars: Verdict and star ratings

Autocar’s philosophy on star ratings is different from others you’ll find. Having gone to the greatest lengths we can both to understand a car, and to assess both its performance and its fitness-for-intended-purpose, we do not aggregate or add up our section scores. Neither do we simply give the very best car in any given class five stars, the very worst none, and filter the rest somewhere in between. It wouldn’t really take a road tester to do that, after all; a spreadsheet would do the job just as well.

Rather, we try to rank cars in a more abstract way that actually means more, and that tells you more about the car about which you’re reading – and that also has more permanence than a five-star score that only lasts until the next class champion comes along.

And so every star rating that Autocar gives actually means something tangible. Their explanations are below:

 - No stars: Inherently dangerous or unsafe. Tragically, irredeemably flawed.

 - Half a star: Appalling. Massively significant failings.

 - One star: Very poor. Fails to meet any accepted class standards.

 - One-and–a-half stars: Poor. Within acceptable class boundaries in a few areas. Still not recommendable.

 - Two stars: Off the pace. Below average in nearly all areas.

 - Two-and-a-half stars: Acceptable. About average in key areas, but disappoints.

 - Three stars: Competent. Above average in some areas, average in others, outstanding in none.

 - Three-and-a-half stars: Good. Competitive in key areas.

 - Four stars: Very good. Very competitive in key areas, competitive in secondary respects.

 - Four-and-a-half stars: Excellent. Near class-leading in key areas and in some ways outstanding.

 - Five stars: Brilliant, unsurpassed. All but flawless.

This approach has its limitations, too – and we know them well. In an increasingly competitive modern car market, we end up with very few five star cars, and even fewer one- and two-star.

But it does allow us the perspective to stand back at the end of the road test process – to survey all that we’ve learned, and not only everything the car does, but everything it is – and then take the broadest view possible on a lasting score. 

Our section scores aren’t arrived at in quite the same way, and if you aggregate them you won’t necessarily arrive at the same final verdict that a car may have actually been given. 

But a century of experience has gone into assessing and pronouncing on new cars in the way that we have come to do it. More empirical data, testing expertise and relevant insight goes into our verdicts today than any other you’ll find. Just as they ought to be in order to be worthy of the very best cars that they address and the trust that you place in them, we believe them to be the very best in the business.