There are lots of things you can do to improve your car's performance - induction kits, upgraded exhausts, different cams, etc. But most of them give relatively modest improvements for what they cost, and/or have significant downsides (long-period cams will inprove power but lose low-down torque.) My general view is that the most cost-effective performance upgrade is to buy a better car. I know this sounds expensive, but you can easily spend thousands of pounds on upgrade components. Buying a better car may well work out cheaper, and also means that the brakes / cooling / supension / etc will be appropriately rated for the performance.

Of course induction kits and uprated exhausts also sound better, which may well be important to you. And I appreciate people like to have a "personalised" car, unique to themselves. But I honestly believe that, long-term, you get more "bang for your buck" by just getting a better car in the first place. (Of course, since I work in the car industry, you may well think "Well he would say that, wouldn't he?" :-) )

You may think it better to buy a low-power car and upgrade it, because the insurance on a high-power car is more expensive. Unfortunately most insurance companies are very wary of modified cars and the premium may be just as much as for the higher power car. You could not tell them, but (a) that would be illegal, and (b) if you crash the car and they spot the upgrade, they may well not pay out.

Many people say that car engines are built "on the cheap" and, with more expensive parts, could be made more powerful. Of course that's right, but that just makes buying a better car so much better value! You could get 150 horsepower from a 1.4, or even a 1.2, if you threw thousands of pounds at it. But for the same money you could get a 2.0 engine that makes 150 bhp in showroom condition, has a lot more torque (so is more driveable), is more reliable, and has a full warranty. Which would you prefer? When you include the cost of fitting the upgraded parts and rolling-road testing to optimise/check the fuel and ignition, it is quite common to be paying upwards of £50 per extra horsepower for a performance mod - and that is pretty close to the extra cost of a better car.

It's not that car makers are necessarily more clever, or better at designing powerful engines, than the aftermarket tuning companies - indeed, the reverse may well be true. But since the car makers can spread the development costs over tens or even hundreds of thousands of sales, it usually works out cheaper for the same quality of engineering. Another major point is that, since 1993, car makers have been forced to fit extremely sophisticated engine management systems to control ignition and fuel, in order to meet emissions limits. A side-effect is that the EMS can be (and is) easily set up to achieve almost ideal fuel and spark for best performance, so the room for improvement that used to exist with crude carburettor and distributor systems is all but eliminated.

It is also worth pointing out that more recent cars - essentially those registered in Europe from 2000 onwards - contain increasingly sophisticated "on-board diagnostic" systems, which may react to a significant performance modification by illuminating the Check Engine Light and/or entering a reduced performance "limp-home" mode.

Additionally, many performance upgrades - removal of the catalyst being the obvious example - result in a very large increase in toxic emissions. For some people this is irrelevant, but readers with a trace of social conscience should think very carefully before going down this route.

Having said all that, there are circumstances when you might want to go down the aftermarket tuning route. This might include racing cars, classic cars (indeed anything built before 1993), and where you already have the biggest-engined version of your chosen car. If so, you could do a lot worse than visit Dave Baker's pages for some facts on tuning (as opposed to the nonsense you will often find on the Net). This isn't any kind of recommendation one way or the other on Dave's services, but certainly his information seems sound.

Aftermarket kits produced with the approval of the original vehicle manufacturer are in general more trustworthy than those that are not, since you can be confident that appropriate safety and durability tests have been carried out, and that the improvements are genuine.

Especially critical for anybody making performance modifications to their car is to understand the difference between engine (flywheel) power, and power measured at the road wheels. Car makers always quote engine power measured at the flywheel, and naturally this is what people want to see increased. But if you have your car "dynoed" on a rolling-road, all that can be measured is the power at the road wheels. A fair amount of power is lost between the engine and the wheels, and so rolling road operators apply a "fiddle factor" to the measured value in order to estimate the flywheel power.

Now, that is all well and good, except that there is no easy guide to what this "fiddle factor" should be, and so it is all too common to over-estimate the "transmission losses" in order to produce an "inflated" flywheel power estimate. Imagine you start with a 100 bhp car and make various performance modifications to it, then decide to have it rolling-road checked. The result might come out as 80 bhp at the wheel, to which the dyno operator will apply the transmission loss estimate and tell you your car is making 115 bhp at the flywheel. Naturally, you are very pleased since your performance mods have added 15 bhp. But...that assumes a transmission loss of around 35%. If you used a more plausible estimate of 20%, you would say that in fact your car is still making the manufacturer's rated power of 100 bhp at the flywheel and so your mods have done nothing at all!

Unfortunately the transmission losses vary from car to car (4WD is much higher, for example) so the only way to be sure is to measure the road wheel power before and after making modifications. Otherwise, it is extremely easy to fool yourself. To make matters worse, some dyno operators may not even quote the "raw" road wheel power figure but will simply present a graph of estimated flywheel power without explicity stating the transmission loss estimates used.

Finally, don't forget that simple test-to-test variation is a percent or two; more if temperatures of air, oil, water etc are not consistent between the "with device" and "without device" tests. Some unscrupulous people will take (say) 5 runs in baseline condition, then 5 with the device fitted, and graph the best "with" result against the worst "without" result. Technically the data is real, but it means you can see maybe a 5% gain in performance from a device that actually does nothing at all!

You can find some good material on this subject here, here and here.

For information on "chipping", and in particular the resistors that are plugged into the air intake temperature sensor connector, see this page. For information on water injection, see here.

Finally, a frequent supposition is that performance modifications will automatically improve fuel economy, at least as long as the extra performance is not used. The logic is that the performance modification somehow gets more power per unit of fuel, so for the same amount of power the fuel used must be reduced. In fact there is very little truth to that argument, as I explain here.