The fleet
Which cars Switzerland kept
Two vehicle censuses, fifteen years apart. From the same model year, three quarters of one model are still registered – and one in eighty of another.
Age is not what decides
You can sort every model line by how much of its fleet is gone. Do that and you get a list of cheap 1990s cars, and what you have measured is age, not care. It only becomes interesting inside a single model year: same country, same fifteen years, same age. The difference is then not small – it is a factor of fifty.
Model years 1990 to 1999
Fleet in 2011 against 2026
Land Rover Defender 110TD4: 75 %. Hyundai Santamo 2000 4x4: 0.6 %.
Still here
- Land Rover Defender 110TD4 2'749 → 2'062 75 %
- Honda S2000 735 → 528 72 %
- BMW Z3 4'110 → 2'556 62 %
- Porsche 911 Carrera 2'617 → 1'579 60 %
- Porsche Boxster 1'161 → 689 59 %
- Jeep Wrangler 2.4 690 → 386 56 %
- Mazda MX-5 1.6 3'461 → 1'921 56 %
- Mercedes-Benz SLK 200 5'280 → 2'829 54 %
Gone
- Hyundai Santamo 2000 4x4 1'588 → 9 0.6 %
- Opel Astra F14 Cvan 1'039 → 6 0.6 %
- Daewoo Nubira 1.6 1'484 → 12 0.8 %
- Kia Pride 1.3 818 → 7 0.9 %
- Opel Sintra A22 2'331 → 20 0.9 %
- Renault Laguna 2.0 Break 2'017 → 20 1.0 %
- Opel Vectra B i500Cvan 8'348 → 108 1.3 %
- Skoda Felicia 1.3 684 → 9 1.3 %
265 model lines from 1990–1999 meet every criterion.
Model years 2000 to 2009
Fleet in 2011 against 2026
Mazda 2 1.5: 84 %. SEAT Cordoba: 2.5 %.
Still here
- Mazda 2 1.5 554 → 466 84 %
- Land Rover Defender110 TD5 1'001 → 761 76 %
- BMW Z4 sDrive23i 608 → 462 76 %
- VW T5 3'087 → 2'342 76 %
- Porsche Boxster 868 → 642 74 %
- Mercedes-Benz SLK 200 K EC 3'500 → 2'558 73 %
- BMW 125i Cabrio 503 → 340 68 %
- VW T5 4Motion 1'725 → 1'154 67 %
Gone
- SEAT Cordoba 609 → 15 2.5 %
- Citroen C5 3.0i V6 2'255 → 79 3.5 %
- Mazda Tribute 2.0 2'106 → 89 4.2 %
- Land Rover Freelander2.5V6 555 → 24 4.3 %
- SEAT Ibiza 3'705 → 162 4.4 %
- Peugeot 307Br 2.0HDI/FAP 1'412 → 62 4.4 %
- Ford Maverick 2.0 1'879 → 83 4.4 %
- Kia Carens 1'325 → 59 4.5 %
363 model lines from 2000–2009 meet every criterion.
Why the obvious analysis is wrong
Three things distort the comparison, and all three are excluded here – otherwise you measure something other than what you think you are measuring.
- Passenger cars only. The census counts everything with a plate; the single largest 2026 entry is a trailer manufacturer.
- Only model lines with an EU type approval. A key based on the model name alone throws a 2015 Sprinter in with a 1995 one, and the line appears to have grown by 69 %.
- No model line still sold new during the window. A Suzuki Jimny, twenty years on sale, otherwise reaches 117 % “survival” – those are arrivals, not care.
- Only lines with at least 500 vehicles at the start. Below that, a single garage moves the percentage.
- And: an observation, not an experiment. Imports, exports and re-registrations all move these numbers.
What this means, and what it does not
The list of survivors is full of cars that are small, cheap to garage and fun to drive. That is not a verdict on build quality: a car stays in the fleet not because it was better made but because somebody decided to keep it. And a low figure does not mean a car was bad – only that nobody saw a reason to hold on to it.
And your car?
Enter your type-approval number and see how many of your model line are still registered in Switzerland.
Look up a vehicle