The cars : MG Metro 6R4 Group B

The MG Metro 6R4 was Austin Rover’s entry into Group B, the controversial rallying category that gave us a series of spectacular cars before being banned at the end of 1986.

What made the MG rather special was its bespoke 3.0-litre V6, the only non-turbo’d car in its category, and its scintillating performance. Sadly, the Metro 6R4 was killed long before it reached its potential in rallying.

MG Metro 6R4 : The giant-killer that never was…


As far as Austin Rover and its parent company BL were concerned, 1981 had marked a sea change in the attitudes of company executives and dealer principals across the land. No longer were they apologetic for the state of their ageing range of cars, but shored up by the warm reception given to the Austin Metro and the fact that new models were in the pipeline, they felt as though there was now light at the end of a very dark and long tunnel.

It was in this climate of optimism that the then Austin Rover Motorsport chief John Davenport hatched a plan. The Triumph TR7 V8 and TR8 rally cars had hung up their competitive boots after the 1980 Lombard RAC Rally and Davenport wanted to replace them with something more in the mould of the ill-fated Ford Escort RS 1700T – a ‘silhouette’ rally car, that employed the classic front engine and rear-wheel-drive transmission package.

Unlike the stubby Ford, Davenport soon decided to follow the lead of Audi and go with a permanent 4WD transmission. Funds were tight (as one can imagine) and it took some persuasion at Board level to obtain funding for the project. Davenport: ‘I managed to persuade our Director Tony Ball to part with £50,000 for a special NEC Show project car.’

Developing the rally Metro

Once the first hurdle was cleared, the next was the development of the car itself – something removed from the traditional art of producing a rally car based on an existing model. Davenport had no qualms about collaborating with the world of Formula 1 in order to get him the type of quick-thinking and focused Engineers in order to get the project off the ground in double quick time. The best team in the world in 1981 was Williams – and its Chief Engineer and Designer was Patrick Head – so, logically, an approach was made.

‘We nipped down the road to see Patrick and said let’s see if we can design a rally car,’ as Davenport put it – of course, it was a easy decision to make as, at the time, the company was sponsoring the Williams team. The decision to base the new competition car on the Metro was also an easy one to make.

In 1982, Patrick Head and Williams Engineer, John Piper, got down to the task of actually building the car – and, as Head himself estimates, he spent approximately forty per cent of his working hours on the MG. Patrick Head recalled: ‘It is clearly very difficult to do F1 on a full-time basis and get involved in any detail with another project in a different area, but I suppose we took on the Metro project through force of circumstances.

But what engine?

‘Austin Rover’s competitions boss John Davenport asked us whether we could shoe-horn a Rover V6 engine – which was literally the 3.5-litre V8 with two cylinders chopped off – into the front of a Metro. Williams GPE agreed to look at the project and soon it became quite clear that, to get the engine in, the driver would virtually have to sit in the back seat. Their team driver Tony Pond came down, tried it for size and immediately said, “there’s no way I could drive this thing quickly through a forest stage – I can’t see the front end from where I’d be sitting.” Clearly, we had to think again.’

Head continued: ‘So we turned the whole thing round with the engine at the back, the gearbox ahead of it with drive to both the front and rear wheels. We presented the concept to Rover and they said “This is great”, pushed the “go” button and off we went with the project. We finished it in about a year, delivered three prototypes to Rover in about November 1981 – six months’ development with Williams’ assistance.’

MG Metro 6R4 in Computervision livery

A new chassis was built from the ground up and it consisted of a floorpan fashioned into a seam-welded tubular chassis – old hat in F1 terms of the day, but still a very effective solution. Williams designed the gearbox, which was produced by its own contractors; the differentials were produced by 4WD experts, Ferguson (of Jensen FF fame). The finished car was delivered to Austin Rover in Cowley as promised in December 1982 – and in-house development began.

Making sense of the Metro

The car arrived in kit-form and it was up to Austin Rover Motorsport to assemble it and test it in anger. For the purposes of comparison, they ran it against a Group A-spec Rover 3500 and an Audi Quattro, the rally car of the time. Even in these early stages of development, the Metro turned in a more than competitive performance despite its under-powered(!) 240bhp development engine. What the first prototype lost in straight line speed, it made up for in agility.

In terms of an power unit, there was little consideration given to using an existing ARG engine (supercharged or turbocharged), but due to what they were likely to be up against, the brave decision was taken to produce a bespoke engine for the new car. Two engines were apparently tried in development prototypes during late 1983: the Honda V6, which was due to appear in the Rover 800 in 1986, and the Rover V8.

Needless to say, Patrick Head preferred something lighter and more compact – and pushed for an all-new V6 loosely based on the Rover V8. From the initial design of a V6 version of the ex-Buick engine, it was Development Engineer Cliff Humphreys at Cowley who managed to move the concept towards reality; making it work beautifully – thanks to a lot of hard work. From there, Stan Johnson and Rob Oldaker at Longbridge completed the transformation.

First public outing

The initial versions of the 6R4 looked almost tame compared with the final, definitive version: As can be seen in this picture, the February 1984 version was based, cosmetically, around the original Metro. The front air dam and sidepods would grow significantly and the rear spoiler had yet to make an appearance. Here, Tony Pond is at the wheel – and he would perform the majority of the test and development driving.
The initial versions of the 6R4 looked almost tame compared with the final, definitive version –  as can be seen in this picture, the February 1984 version was based, cosmetically, around the original Metro. The front air dam and sidepods would grow significantly and the rear spoiler had yet to make an appearance. Here, Tony Pond is at the wheel – and he would perform the majority of the test and development driving

The MG Metro 6R4 (ie: 6-cyclinder, Rally, 4-wheel-drive) made its first public appearance at the end of February 1984 in a hastily prepared press launch at the Excelsior Hotel, London Airport. The reason for this was that press speculation was mounting – and, in order to undergo a full development programme, Austin Rover could do without the added hassle of playing hide and seek with scoop photographers.

At the launch, Tony Pond drove the car onto the stage and said of it, ‘It’s fantastic! You go into a corner deep, get the back end out under braking – and then drift through with the power on. If you go in too deep you simply use more power.’

The rest of 1984 was used to finalise the 6R4: a final engine needed to be developed and the aero package also needed more work.

The cut-and-shut Rover V8 engine of the development car was replaced by the specially-designed and built four-cam, 24-valve, normally aspirated 3.0-litre V6. The power output was quoted at 410bhp at 9000rpm at the car’s official launch in May, 1985 – and it was promised that this would soon be improved upon. The aero package was modified and took the rulebook to the extreme, stretching the credibility of the ‘silhouette racer’ ideal to breaking point. However, that did reflect the-then current Group B thinking also followed by Peugeot 205T16 and Lancia Delta S4 – looking back at it now, it is easy to see that the cars were becoming completely over the top for the task in hand.

Entering competition


Engine Capacity: 2991cc, Bore: 92mm, Stroke: 75mm
Valve Gear: DOHC, 4 valves per cylinder, toothed belt camshaft drive
Compression ratio: 12:1, Lucas mapped electronic fuel injection
Configuration: Longways/mid mounted, 6 cylinders in 90 deg, V, dry liners, 4 main bearings, water cooled
Maximum Power: 410bhp at 9000rpm, Maximum torque: 270lb ft at 6500rpm

Austin Rover Motorsport announced at the time that the MG Metro 6R4 would be entered in the 1986 World Rally Championship – and would be competing against such definitive cars as the Audi quattro S1, Lancia Delta S4 and Ford RS200. Group B rallying had entered a truly exciting – and, as we would soon see, terrifying – era.

Meanwhile, Autocar magazine strapped its timing gear to the Clubman (250bhp) and International (410bhp) versions of the 6R4 and produced some simply awesome acceleration figures (The figures for the International version are quoted):

0-30mph 0-40mph 0-50mph 0-60mph 0-70mph 0-80mph 0-90mph 0-100mph 0-110mph
1.2s 1.7s 2.4s 3.2s 4.0s 5.2s 6.5s 8.2s 10.0s

The magazine came away impressed (who wouldn’t be after turning in those figures) and it was John Davenport who, when asked by Autocar whether the MG Metro 6R4 would win rallies, summed up the company’s optimism about the car’s chances: ‘Yes. I see no reason why it shouldn’t. I think the car is quite capable of doing it, and I think my drivers are more than up to the job.’

Sadly, it was not to be: Henri Toivinen was involved in a fatal collision, when his car went over a cliff on the Corsican Rally. The FISA responded by ruling that Group B rally cars were simply too fast and therefore dangerous to drive and, as a result, banned them from competition effective from the end of that season. The international career of these Group B cars in international rallying was now over – and the book on the promising 6R4 was closed prematurely.

Great potential, unrealised

Austin Rover Motorsport gave up on rallying at this point – and it would not be until 2001 that we would see another factory-backed MG rally effort. The car lived on in a couple of ways: road versions were eventually sold off at a bargain £13,000 a piece, many entering private hands to be used to devastating effect in rallycross events across Europe.

The engine design was then sold to TWR (Tom Walkinshaw Racing) and, after some development work on the cambelts and plenums, the engine would make a re-appearance; initially in 3.5-litre form in the Group C Jaguar racing cars, but then in the back of the sensational 217mph Jaguar XJ220.

The works MG 6R4 rally car caught during testing.
The works MG 6R4 rally car caught during testing

In car video


With thanks to Ken Smith for the Williams GPE information, Dan Ellmore, David Sims, Nicky Lindon and Malcolm Boote.

Keith Adams


  1. With the finances in AR always in dire straights, how long would they have stayed in WRC, i think that they could have gone on to become a major player in the world of rallying, but we will never know, shame really as the 6R4 and RS200 were great rally cars for there time.

  2. Was the engine counterweight in any way to try and eliminate the balance vibrations from lopping 2 cylinders off a V8?
    Obviously they did a fine job if it even found it’s way into the halo Jag!

    Still popular recently in Rallycross, where Group B lives on!

  3. Given MG’s just announced return to racing isn’t it about time someone did a book on the history of Austin/MG/Rover/etc on/off involvements in motorsport (of all sorts) over the last 100 years or so?

  4. Will M. As the article says “The cut-and-shut Rover V8 engine of the development car was replaced by the specially designed and built four-cam, 24-valve, normally aspirated 3-litre V6.” This bespoke engine was quite different and designed to be a V6 from the beginning and not a warmed over V8 minus 2 cylinders.

  5. What’s under the bulgy bonnet of the Computavision-car at the top of the page? It seems the other cars make do with an almost standard bonnet.
    And I always thought the 6R4 was longer than the normal Metro but in the (development) pictures I see more or less standard bodies.

  6. what a lovely little and ugly nice car this is… i heard him years ago roar on the zandvoort circuit… nice…

  7. Considering the T-Series Turbo is said to be capable of 800 bhp, had Group B not been cancelled would the Metro 6R4 have eventually evolved into a Metro “4R4” with a heavily modified M-Series Turbo engine?

  8. @9, Nate,

    A Metro with 4wd and 800bhp- you’d need your own private airfield with very wide runoffs- and cahoneys the size of watermelons to drive that thing!

  9. 10 – Chris Baglin

    Obviously the bhp would be much lower for rallying purposes (Pikes Peak is another matter…) though it does makes you wonder since the most powerful Group B car was said to have almost reached up to 600 bhp (with the final versions of the Group B Audi Quattro rated at 591 bhp) prior to the series being cancelled.

    A successful Metro “4R4” could have potentially been used to further promote the M-Series Turbo by giving it a motorsport pedigree (albeit in a very heavily modified form), even if little consideration was given to using an existing ARG engine in place of a bespoke 6-cylinder in real-life.


  11. Good to see a photo of one of the development 6R4’s on the 1984 International Welsh Rally in the Ternco livery.

    Some will remember that this was used for the TV drama “The Winning Streak” driven by the character David Savage.

    I was a spectator on the ’84 Welsh and have very fond memories of the 6R4’s during the golden age of rallying which was of course Group B.

    Although Tony Pond is no longer with us he will always be remembered for his exploits in the 6R4 on the RAC Rally, especially in 1985 when he finished 3rd overall.

    Great memories!

  12. @14, It would not have conventional throws on the crank- like a DFV it would have a flat plane crank an would sound almost like a inline engine.

    At 9000 rpm it would have obliterated itself with a regular crankshaft,

  13. God these were mental….
    Sounded great but Dam ugly ! 🙂
    I watched them in the late 80’s charging through forest near were I live, along with the Quattro’s & 205’s & Delta’s etc
    I soon realised that Rally driver’s are just plain bonkers!

  14. I believe that the ARV6 engine was quite successful in the World Sportscar Championship C2 Ecurie Ecosse car driven by Ray Mallock.

  15. I believe it was quite a “shoehorn job” to get all the mechanical bits into this size of body. Had Group B survived a bit longer, it would have made sense to put these mechanicals into a Maestro body. Now there was a car which needed an image boost, especially to the under-90’s age group!
    This would of course have been a reversal of the Big Bertha to Baby Bertha progression at Vauxhall.

  16. @18

    Funny you mention Vauxhall, some of these prototype 6R4 pictures look very Nova-ish.

    Perhaps if the MG version had kept some of that style it would’ve been as popular with the youth. The basic straight lined hatchback shape of both cars was quite similar.

    • Running a 90′ V6 is easy. Buick did it for years in two configurations, initially with a shared crankpin and odd-fire ignition, then as a separate crankpin model and an even-fire ignition set up. Anyone wanting to build a special V6 for a high power alternative to the ARG 24 valve engine can buy Buick V6 alloy blocks with up to 4″ bore, and big valve alloy heads which flow high numbers, and of course there is significant experience of turbocharging Buick V6s to give huge power ratings. TA Performance is the first call for high-perf Buick parts. For those doing it on the cheap, a good starting place would be a GN spec iron engine. There is no cheap alternative for the transmission I fear ……..

  17. A few thoughts-
    credit for the V64V engine design goes to David Wood.
    The bulgy bonnet came about, sadly, because it was found necessary to increase the front suspension travel, and the only way to do that was to push the strut tops upwards.
    The 6R4 was extremely successful in UK National Rallying after the Group B ban, despite regular attempts to handicap it out of contention !

    When the 200 off homologation batch had been built at Longbridge, the cars were offered (in basic Clubman tune) at £13K each. I’m just one of the many people who kick themselves now for not snapping one up !

  18. “the car’s official launch in May, 1985….Group B….banned at the end of 1986”

    Can’t believe it was so short lived!

  19. I remember “The Winning Streak” being broadcast about a rally team. It starred the first incarnation of the 6R4. I also saw the first incarnation in action at Brecon, once. I have eh photos and the memories of it.

    • If i remember correctly it also had Dinah Sheridan and The Guy who went on to play Inspector Wexford as the owners of the Dealership chain, also if my failing memory serves me correct Wexford drove a stunning golden brown R3500 V8 – there was also another series around the same time, and i am sure that it was about a small car company that was producing a new hatchback, Maestro, with different face, and they were trying to sell it to the police and it failed on the day of testing for the force, however i can not find any reference to it, or remember the name… I would love to see both shows again though,

      • Just checked, wish i had to start with… LOL, it was the great actor Leslie Sands that played the lead actor within the family, but he does look like Wexford LOL….. told you my memory was failing.

  20. Bernie Marcus was the ARG Motorsport aerodynamics who developed the package of changes from where Williams finished. It was said that the airflow was found to be coming out of the short sidepods rather than going in, hence there extension into the doors.

  21. Also read elsewhere David Wood was the one who designed the specially-built four-cam, 24-valve, normally aspirated 3-litre V6 used in the MG Metro 6R4 and reputedly used some of the engine architecture of the Cosworth DFV V8.

    Along with the Jaguar XJ220, the engine would also find its way into the experimental 300 hp Lotus SID (Structures, Isolated and Dynamics Research) project of 1992 as well as spawned the basis of a 3.5-4-litre V8 developed by Cosworth Engineering and used in the 1994 Bentley Java concept (despite claims of featuring a BMW-sourced V8 unless the latter was later used in the one-offs commissioned for the Sultan of Brunei). Also seem to recall the engine (or possibly even some variation of the DFV) being used in other cars both for motorsport along with cars that never reached production.

    In theory despite the engine’s convoluted history, it could have potentially formed the basis of a rather sophisticated V6/V8 engine family for Ford Europe in other circumstances with the 3498cc displacement of the V6 Twin-Turbo used in the Jaguar XJ220 equating to a 4664cc V8 (in contrast to the Essex/Cologne V6s and Windsor V8).

    Though the following would have likely had less success compared to the MG Metro 6R4, the company earlier on (albeit in ideal circumstances) could have looked to the mid-engined 2WD 4-cylinder turbocharged Renault 5 Turbo Group B competitor and developed a lower-tier RWD turbocharged version of the MG Metro along similar lines.The Group B class the Renault 5 Turbo was placed in had a maximum displacement limit of up to 1428cc and weight limit of 820kg.

    • Sadly, you are misinformed regarding theJava. The powertrain and chassis for the Java was carryover 540, as was most of the electrical architecture. (I had a very close association with the project.)

      • Understand, have also read Cosworth developed a mock-up 3.5-litre V8 engine that was unveiled alongside the Java concept (which did carryover the powertrain and chassis of the 5-Series) though never came to fruition. From when Vickers owned both Cosworth and Rolls-Royce/Bentley.

        Some say the unbuilt road-going twin-turbo Cosworth V8 was coded LF, with 75-degrees and derived from a MBX V6 that was to be part of a family of related road-going engines yet not an adapted racing engine, while others claim it was essentially a V8 version of the twin-turbo V6 used in the XJ220. At least what was mentioned in Graham Robson’s book on Cosworth as well as Graham Hull’s Inside the Rolls-Royce and Bentley Styling Department.

    • By the way…….the only ‘one-off’ Java was the show car – launched in green, later repainted red. There were 18 others built for the Sultan…..6 drop heads, 6 coupes, and 6 shooting brakes.

      • A bit surprised by the Sultan not having a 4-door saloon version commissioned, others can disagree though find the Java concept to be a very attractive design that unfortunately did not reach production (the same goes with the Rolls-Royce SX proposal of the early 1980s).

        Know an attempt was made at carrying over the Java’s styling onto the E39 chassis that apparently did not translate as well before the whole project was shelved, a shame really given the E39 was a very well regarded car (the less said about the Continental GT the better).

    • The basis for the R64V engine was indeed the Rover V8. The project was called RedCap (Reduced Capacity) and was a Triumph project to replace the PE146/PE166 engines in SD1. At around this time the ‘O’ Series was being considered as a replacement for the Triumph Slant engines in TR7 and also for MGB. Due to the appalling emissions from the ‘bowl in piston’ ‘O’ and its poor idle and NVH qualities it was evident that it could not meet the Federal emissions standards of the time. This led to the 4 Valve ‘O’ Series TR7 engine being developed at Canley. The longer term aim was also to sell the SD1 into the North American market and the conclusion was therefore that a RedCap engine should also have 4 Valve heads and that the Rover V8 should also receive a version of these heads.

      The first RedCap prototype engine went into an SD1. The engine was a full blown V8 with the front cylinders intact but not containing any pistons. The crankshaft for this engine was designed by Hesketh Engineering at Towcester and produced by Farndon of Bayton Road, Exhall. It ran on triple IDA Weber carburettors and retained the standard V8 Distributor with 2 cylinders disabled. This car was lent to Cliff Humpheries at Abingdon for evaluation of the engine. The 4 valve heads were initially designed jointly by two Engineers in the Triumph Engine DO. These first heads bore an uncanny resemblance to the 4V 0 TR7 heads. Later, as Nate says, David Wood from Cosworth came on the scene and took over design responsibility. This led to a complete block, head, crank and valvegear drive redesign. The engines were machined at the Triumph engine facility at Capmartin Road, Radford, Capmartin Road was a rear extension of the Jaguar Radford plant but operated by Triumph.

      Around this time I left the Triumph Fletch North Engine development section as we were now working on durability work for Longbridge A and O series engines and also doing durability on the E series for the first LM10’s and early work on the S series ‘reversed’ engines for the later LM11. If we thought the O Series (nicknamed the ‘O Deries’ at Canley) was bad from a combustion point of view the E was diabolical and the S maintained the reverse splayed valve system!

      I moved to the Lucas PI lab in Farm Street, Aston, Birmingham. And then got involved in the S twin cam engine which ARG had no money to develop so they asked Lucas to do the work as an unpaid ‘Favour’! The twin cam version of the S was a revelation, after some work to improve the calibration ‘in vehicle’ it was a stunning performer. It would rev and was quite capable of 120 mph in a 216 as a certain BMW driver knows when he tried to race me on the M6 – fortunately for me I knew the law was not averse to siting on the southbound slip road at Junction 1 waiting to catch speeders – he didn’t!

      Later I moved into the Lucas Racing Department to assist with calibration of the Micos PI systems and got involved with the calibration of the R64V engine. It was a tricky engine to calibrate from a system point of view because of the odd fire. We ran the engine as if it was a 24 cylinder using a 12 bladed crank trigger wheel (this being in the days of ‘time based’ control systems rather than an ‘angle based’ system as common today). This meant ‘masking’ of 18 of the ’24’ cylinders the system was seeing, to fire the remaining ‘correct’ 6 cylinders.

      The R64V engine was very fragile from a torsional vibration resonance at 7500 rpm. This was so destructive that it caused the cam belts to break or jump off, valves springs to surge uncontrollably resulting in cotters jumping out and valves dropping, or just crankshaft breakage which could be spectacular. I think it was the Eccurie Ecosse entered R64V engined car that was leading the C2 class at Le Mans in 1988 which failed with the crank breaking after they increased the rev limiter from 7,300 to 7,600 for the last 2 hours of the race – I remember it as a White car with a Japanese sponsor (Curtain Dreams!?).

      It could have been such a good engine had they either made it as a 60 degree or as an offset pin crank to give it an even firing order.

      • Thanks for the background Martin

        It seems the R64V because of its Rover V8 roots via the 90-degree layout at the end of the day would have still benefited from the type of updates the Buick V6 later received in 1977 and 1988 to remain in production until 2008, can assume the same limitations would have also applied to the BL Australia developed 3.3-litre V6 version of the 4.4-litre Rover V8 let alone the entire idea of an all-alloy V6 version of the Rover V8 in general.

        A few questions:

        – The limited information available in the James Taylor books regarding the 2.8-litre RedCap V8 project either mention it as being conceived chronologically in the early-70s with the P10 saloon being one projected application at one point or as part of an early-80s Land Rover/BL project called Adventurer that was intended to develop a common structure for Land Rover, Range Rover and even Freight Rover-type models.

        It is known what the exact bore / stroke and projected output of the RedCap V8 was planned to be and what influence (if any) if it had on TVR later introducing the short-stroke supercharged TVR 2-litre V8S for the Italian market?

        – Do any approximate figures exist for the Twin-Cam S-Series engine? Additionally as the Twin-Cam was fitted to the Rover 216, how did the engine compare to the 120-130 hp Honda D16 powered Rover 216/416 that was likely its internal benchmark or the later 111 hp 1.6 K-Series that ended up superseding the Twin-Cam S-Series?

3 Trackbacks / Pingbacks

  1. Four-Links – Roman holiday, Borgward’s Traumwagen, vintage racing games | Hemmings Daily
  2. Four-Links – Roman holiday, Borgward’s Traumwagen, vintage racing games, Darrin’s Darrin | 3Buzz Now
  3. Four-Links – Roman holiday, Borgward’s Traumwagen, vintage racing games, Darrin’s Darrin - The Classic Cars Post

Leave a Reply

Your email address will not be published.


This site uses Akismet to reduce spam. Learn how your comment data is processed.