From the Williams six-wheeler banned before it could race to the Honda that died with its designer, these are the Formula 1 cars that were tested but never made their competitive debut
Jacques Laffite in the cockpit of Williams' six-wheeled FW08B in 1982
Formula 1 has always operated on the principle that what is not explicitly forbidden is permitted.
This has led to some of the most creative, outlandish and legally questionable innovations in the history of motor sport.
But for every idea that makes it to the grid and survives long enough to win races, there are dozens that are tested in secret, judged too effective, too expensive, too dangerous, or simply too inconvenient for everyone else – and promptly buried.
Some never made it beyond the circuit gates. Others were banned before they could prove themselves in competition. A few were destroyed by fire, funding, or sheer bad luck.
These are their stories.
Cosworth 4WD F1 car (1969)
In 1969, four-wheel drive was the future. Or so everyone thought.
Engine manufacturer Cosworth built its own complete 4WD Formula 1 car that year, reasoning that more driven wheels would mean better traction and faster acceleration.
Jackie Stewart tested it and was polite but unconvinced: “The car’s so heavy on the front, you turn into a corner and the whole thing starts driving you.”
Bruce McLaren built a 4WD car (the M9A) and described driving it as “trying to write your signature with someone constantly jogging your elbow.”
Lotus and Matra also built 4WD cars. None of them worked particularly well.
The problem was weight. Four-wheel drive systems added significant mass to the front of the car, which destroyed the handling balance and made the cars sluggish through corners despite the theoretical traction advantage.
And then aerodynamic downforce arrived, which provided more than adequate traction without any of the weight penalty.
The Cosworth car, the first and only car the company ever built, was tested but never raced. The McLaren M9A raced once at Silverstone, driven by Derek Bell, and retired with suspension failure. The experiment was quietly abandoned.
Four-wheel drive has never returned to Formula 1, and almost certainly never will.
Lotus 88 twin-chassis (1981)
Colin Chapman had a problem. Ground effect – the aerodynamic innovation that had dominated Formula 1 since 1978 – was about to be neutered.
The FIA, alarmed by cornering speeds and concerned about driver safety, had banned the sliding skirts that sealed the gap between the car and the track surface, making ground effect work.
Without the skirts, ground effect would be significantly less effective.
Chapman’s solution was audacious even by his standards: build a car with two chassis, one inside the other.
The Lotus 88 consisted of an inner chassis that carried the driver, engine, and suspension, and an outer chassis that was essentially the aerodynamic bodywork, separately sprung so it could sink to the ground at speed and restore ground effect without technically violating the ban on skirts.
The driver sat in the inner chassis; the outer one moved independently around him. It was brilliant, outrageous, and – according to multiple technical experts and even some FIA appeal boards – entirely legal.
But the car never raced.
The 88 passed scrutineering at the 1981 United States Grand Prix West in Long Beach, practised, and was then black-flagged after rival teams protested. The same thing happened in Brazil. And again in Argentina.
Chapman appealed repeatedly. Expert witnesses testified that the car complied with the regulations. Appeal boards found in his favour. None of it mattered. The politics were insurmountable.
Chapman was so disillusioned by the rejection that many believe he never saw Formula 1 in quite the same light again. He died the following year, aged 54.
The Lotus 88 sits in the Team Lotus collection, one of the great “what ifs” of the sport.
Williams FW08B (1982)
While Cosworth was pursuing four-wheel drive, Williams was taking the opposite approach: keep the front wheels undriven, but add two extra driven wheels at the back.
The Williams FW08B was the second of the team’s six-wheeled experiments, following the FW07D tested by world champion Alan Jones in 1981.
The concept was straightforward: four driven rear wheels would provide superior traction out of slow corners without the weight penalty and handling complications of a 4WD system.
Alan Jones flew back from his final Formula 1 victory in Las Vegas to test the FW07D prototype at Donington. After that, the FW08B was built – the car intended to contest the 1982 season.
As this week in Barcelona has shown, testing throws up some interesting sights – none more so than the six-wheel wonders of the 1980s Keke Rosberg’s championship-winning Williams of 1982 was…
By
Paul Fearnley
Rosberg lapped Paul Ricard in 1min 04.3sec during testing, a widely reported quick time. The car was fast, the drivers liked it, and Williams was preparing to race it.
Then someone in a FOCA meeting said it would drive up costs and cause chaos during pitstops.
The FIA promptly limited the number of wheels for all cars to four, of which only two could be driven. The FW08B was effectively outlawed before it could race.
Only one FW08B chassis was ever built at Williams’ Didcot headquarters. The team was on the verge of manufacturing more gearbox internals and suspension components when the ban was announced. That was the end of the road.
The car has since been restored by Williams and occasionally appears at historic events.
DAMS GD-01 (1995)
The DAMS GD-01 was a victim of timing, politics, and bad luck in roughly equal measure.
The French team DAMS – successful in Formula 3000 – collaborated with Reynard to build a Formula 1 car from 1994 to 1995 with the intention of entering the championship.
The car was extremely conservative by design: ample cooling, unadventurous aerodynamics, and an off-the-shelf Cosworth V8 engine. The plan was to learn quietly, avoid embarrassment, and build from there.
The problem was that Formula 1’s technical regulations were in chaos in 1994 and 1995 following the deaths of Ayrton Senna and Roland Ratzenberger at Imola.
Teams were struggling to interpret the new rules, and DAMS — without the resources of an established outfit — found itself designing to a moving target.
By the time the car was ready for testing at Paul Ricard in October 1995, driven by Jan Lammers and Erik Comas, it was clear the performance simply wasn’t there.
A deal with the ailing Larrousse team to run the car in 1995 fell through. The car was not fast enough, and without significant additional funding to develop it further, DAMS abandoned the project in November 1995.
The GD-01 never raced.
The team returned to Formula 3000 and later became one of the most successful outfits in GP2 and Formula 2.
Dome F105 (1996)
The Dome F105 is probably the saddest story on this list.
The Japanese constructor Dome — experienced in sports cars but new to Formula 1 — built the F105 powered by a Mugen Honda engine and launched it at the Spiral building in Tokyo in March 1996.
The plan was to run it as a privateer entry, learn the ropes, and establish a foothold in the championship.
Marco Apicella, Naoki Hattori, and Shinji Nakano were selected as test drivers. Testing began at Mine circuit and Suzuka.
A Suzuka test ended when an oil leak caused a major fire that extensively damaged the only chassis.
With limited budget and sponsorship, Dome couldn’t realistically recover the programme, and later Suzuka comparison laps (best quoted 1min 46.270sec) suggested the car was off the 107% pace against 1996 benchmarks.
Dome returned to sports cars and never attempted Formula 1 again.
Honda RA099 (1999)
Of all the cars on this list, the Honda RA099 is the one that hurts the most.
Honda had last competed as a constructor in the 1960s, withdrawing after the death of Jo Schlesser at the 1968 French Grand Prix. It had returned as an engine supplier in 1983 and enjoyed enormous success with Williams and McLaren through the late 1980s and early 1990s.
By 1998, it was considering a return as a full constructor.
Honda hired Harvey Postlethwaite, the brilliant designer who had worked at Ferrari and Tyrrell, as technical director, and commissioned Dallara to build the car.
The RA099 completed its first laps in December 1998 at Varano de’ Melegari in Italy, and then joined the 1999 winter tests alongside the established teams.
Everything was in place for a full-scale assault on the 2000 championship
The car was immediately impressive. At Jerez in January 1999, Jos Verstappen set the fastest time, beating Jacques Villeneuve’s BAR by 1.2 seconds.
The next day, Verstappen was fastest again. And the day after that. The RA099 was quick, reliable, and clearly competitive with midfield teams despite being a prototype built in a few months by a skeleton crew.
Rival teams complained to Bernie Ecclestone that Honda was testing alongside them, and the Japanese manufacturer was told to organise its own test programme separately.
Honda’s board of directors began discussing a budget of £200 million per year for five years to support the Formula 1 team and modernise the Bracknell factory. Everything was in place for a full-scale assault on the 2000 championship.
And then, on 15 April 1999, during a test session at the Circuit de Barcelona-Catalunya, Postlethwaite suffered a fatal heart attack. He was 55 years old.
Honda attempted to continue without him. Rupert Manwaring took interim management and affirmed that the team was determined to continue the test programme.
Mika Salo signed a test driver contract to join Verstappen from June onwards. But Honda’s board vacillated.
Ron Dennis was scathing about the manufacturer’s hesitation, calling its behaviour “unprofessional” and noting that Honda’s reservation of the 12th and final Formula 1 grid slot had forced Zakspeed to abandon its planned return.
In the end, Honda pulled the plug.
Instead of entering as a constructor, it agreed to supply engines to BAR and Jordan for 2000. The RA099 was shelved.
McLaren MP4/18 (2003)
The McLaren MP4/18 is arguably one of the most famous cars never to race in Formula 1 history.
Designed by Adrian Newey as a radical departure from its predecessor, it incorporated ideas years ahead of their time, including concepts that would later form the basis of the blown diffuser used on the Red Bull RB7.
There was just one problem: it kept catching fire.
According to former McLaren mechanic Marc Priestley, after only the first test session, the car came back from every single run on fire.
The packaging was so tight, the aerodynamics so aggressive, that the engine and gearbox simply could not be cooled adequately.
Ron Dennis had predicted the car would represent a “quantum leap” in performance. Instead, it became a nightmare.
McLaren spent the entire 2003 season attempting to fix it while racing the previous year’s car – the MP4-17D – as a stopgap.
Test driver Alex Wurz eventually refused to drive the MP4/18 anymore. Kimi Räikkönen, racing the outdated 17D, fought for the championship anyway and lost to Michael Schumacher by just two points.
Had McLaren simply developed the 17D properly instead of obsessing over the troubled 18, Räikkönen would almost certainly have won the title.
The MP4/18 never turned a wheel in anger. Its successor, the MP4-19, inherited many of its problems and was also dreadful.
McLaren would not win another race until 2005.
Bonus entry: Williams FW15C with CVT transmission (1993)
The FW15C itself won the 1993 championship, so this is technically a bonus entry – it’s the CVT variant that never raced, rather than the car itself.
David Coulthard knew something was different the moment he fired up the Williams FW15C at Pembrey in July 1993. The engine sound was completely flat. No rise and fall through the corners.
No crescendo down the straights. Just a single mechanical whine that barely changed pitch regardless of what he did with the throttle.
The reason was simple: the car he was driving didn’t have a traditional gearbox. Instead, it had a continuously variable transmission – a CVT – supplied by the Dutch truck manufacturer DAF.
Instead of fixed gear ratios, the system used a belt running between two variable-diameter pulleys, allowing the engine to sit at its optimum rev range constantly while the car accelerated.
The result was deeply, profoundly unnatural to listen to, and devastating in terms of laptime.
Reports vary on exactly how much faster the CVT Williams was compared to the conventional car, but estimates range from one to three seconds per lap depending on the circuit.
Given that the FW15C with a conventional gearbox was already dominant enough to win the 1993 championship by a mile, the CVT version would have been untouchable.
The FIA understood this perfectly well. Before the car could race, it mandated that all Formula 1 cars must have between four and seven forward gears, and then added a specific sub-clause banning CVT just to be absolutely certain.
Coulthard tested it on a wet day at a Welsh airfield circuit, and Alain Menu drove it once at Estoril. That was the sum total of its running.
