How we test
The whole method fits on one page, which is deliberate. If you cannot repeat a measurement you should not publish it, and if a reader cannot repeat it in a dealership it is not much use to them either.
A second-hand car being looked over in daylight, bonnet up, a torch pointed into the engine bay from a low angle.
The five steps, in order
Set the car up before anything else
Cold tyre pressures against the door placard, read with our own dial gauge, never the forecourt hose, which in our experience reads up to 0.3 bar out. Fuel taken brim to brim at the same pump on the same side. Boot emptied and the spare checked for presence and pressure. Odometer and trip computer photographed. If the car arrives warm, we wait.
Measure the car standing still
Seat base length, cushion angle, H-point height off the tar, door aperture height at the B-pillar, backrest-to-wheel reach, rear knee room with the front seat at 1.78 m, rear cushion height, boot lip height and boot aperture at its narrowest. Boot volume by 20-litre crate as well as by the published litre figure, because the shape matters more than the number.
Drive the loop in the same order
Forty-six kilometres through Westdene, Brandwag and Universitas in the afternoon school window. Then the N1 south to Trompsburg and back at a true 120 km/h, with noise read at 60, 100 and 120 on the same 4 km. Then 11 km of district gravel toward Soutpan at 60 and again at 80. Rattles counted on the return leg with the aircon off.
Load it and measure again
Ride height from wheel centre to arch lip at all four corners, empty and then with the load the car is meant to carry. A rear drop over about 40 mm is marked. We check the payload figure against kerb weight and gross vehicle mass and state plainly whether the car can legally carry what its seat count implies, because many cannot.
Write the arithmetic down, then the opinion
Fuel in litres per hundred kilometres for town and open road, blended for 20 000 km a year at the current pump price. Tyres at a realistic local life. Service plan or maintenance plan read line by line, with the severe-duty interval noted. Only after all of that do we write the two or three sentences that say what we think.
Two honest limits. We measure one example of a car, and build variation exists, so a rattle we found may not be in yours. And a sound meter on an armrest is not a laboratory: treat our decibel figures as comparable with each other and not with anybody else's.