For a really big improvement in fuel consumption, unfortunately the only answer is to choose a more economical vehicle. Obviously it is expensive (in both money and energy) to build a new car, so it makes little sense to scrap an existing vehicle simply in order to replace it with a more economical one. But if you are changing your vehicle anyway, you can obtain a very large reduction in your fuel bill through careful choice.

The most important decision is the fuel type. Diesel vehicles are typically 20% more economical than their petrol equivalents, due to their fundamentally more efficient combustion process, and should always be the first choice for anyone to whom economy is important. Modern diesels are also quiet, powerful and generally far more appealing than those of 10 or 20 years ago. However, they are considerably more expensive to make (due to the higher pressures involved and the precision manufacturing of the injectors), so you need to decide if the extra purchase price would be covered by the fuel savings.

There is of course much debate over the toxic emissions from diesel engines. Certainly the mix of gases is different to that from a petrol engine - generally speaking, more particulates and NOx but less hydrocarbons and CO. Whether or not the diesel is worse is very much an open question, and may depend on the particular environmental conditions in the local area (temperature, etc). Emissions limits for both petrol and diesel engines are being steadily lowered in response to current thinking on the toxicity of specific emissions components; I am not, however, an expert on the health effects of toxic pollutants so cannot offer any more advice on this topic.

"Alternative" fuels such as LPG are generally only cheaper because of lower taxation rather than any inherent economy advantage, and (in my opinion) should only be considered if the extra purchase cost can be recouped in a small number of years. There is after all no guarantee that these fuels will always have lower taxation.

On the other hand, diesel engines are considerably heavier and will never have the same quick response as a petrol (due to the extra weight of the reciprocating parts). As a result many people (including myself) still prefer petrol power, and there is a wide range of technologies on offer to improve the economy of petrol engines:

• variable valve timing (VVT, Vanos, VTEC, etc)
• variable turbulence
• exhaust gas recirculation
• direct-injection (Ford SCi, Audi FSI, etc)
• turbocharging
• friction-reducing measures
• cylinder shut-off (GM "Displacement on Demand", etc)
• engine shut-off at idle (Citroen "Stop & Start", etc)

One principle applicable to both diesel and petrol is downsizing - if you can get more power and torque from a given size of engine, then you can use a smaller engine for the same performance, and so improve economy. So for example if you make a 1.4 litre engine as powerful as a typical 1.6, then it will generally give better economy than the 1.6. Technologies such as turbocharging and VVT are fundamentally performance improvements, but can give better economy as a side-effect.

Another important aspect is the transmission technology. A conventional manual gearbox is almost always the most economical choice, while a "traditional" torque-converter automatic may use as much as 10% more fuel. That does however assume that the manual is always driven in the most economical way; in the real world, the automatic's disadvantage may be considerably less.

For those who prefer automatics (and in today's crowded traffic, there are many good reasons to do so) there are a number of alternative gearbox technologies:

• Continuously variable transmissions (Audi Multitronic, CVT) - these use a belt moving over variable-sized pulleys to achieve a "stepless" change of ratios. These tend to give good economy and smoothness, but the engine speed varies in a way that can be disconcerting to some drivers.

• Automated manual transmissions (Citroen Sensodrive, Ford Durashift EST) - these are conventional gearboxes with computer control of the clutch and the gearchange mechanism. In economy and performance terms they are identical to manuals, but some people find the gearchange-pause during accelerations unpleasant. In truth they are generally quicker and smoother at changing than all but the best drivers of manual gearboxes, but because you are not making the change yourself it is much more perceptible. Many motoring journalists hate these gearboxes, but in my experience they are perfectly acceptable.

• Audi's DSG twin-clutch gearbox seems to be pretty much the best of all possible worlds, and probably represents the future for automatic gearbox technology

Whichever gearbox technology you choose, for the best fuel consumption you want a large "spread" of ratios - a "short" bottom gear for good acceleration, and a "tall" top gear for economical cruising. In most cases this means more ratios - going up from 4 to 6 speeds gives significantly better economy and performance; somewhere around 7 is generally believed to be the theoretical optimum for passenger cars.

Interestingly, part of the reason why diesel cars give good economy is that their high torque at moderate speeds allows them to have very "tall" gearing while still giving good driveability. This means the engine is turning more slowly in normal driving, which in turn reduces friction losses.

Hybrid vehicles are in the news a lot at the moment. The principle here is to supplement the conventional petrol engine with an electric motor for low-speed town driving and added acceleration; the motor also captures energy that would otherwise be lost during braking. Certainly in town driving hybrids are excellent, giving good fuel economy and also a smooth and responsive drive. On a steady cruise, however, the electric motor is doing nothing and so the vehicle is no more economical than a "non-hybrid". Some motorists who drive mostly on long motorway journeys have been quite disappointed by the economy of their hybrid vehicles and would arguably be better off with a diesel. (See also this recent study.) A hybrid may well be your best choice, but think about your typical driving pattern first.

The type of vehicle is also a very significant factor. Larger, heavier vehicles will always use more fuel than smaller, lighter, ones - all other things being equal. If you want a large, heavy vehicle such as an off-roader (4x4, SUV), but also want good fuel economy, you will have to pay extra for sophisticated engine technology. Certainly some large European diesel "executive cars" such as the Audi A8 or Mercedes E-class have astonishingly good fuel consumption, as good as you would expect from a much smaller and less luxurious car with a "low-tech" petrol engine.

(There is a lot of talk at the moment about possible extra taxes or restrictions on 4x4 / SUV vehicles, mainly because they are seen as environmentally unfriendly. Personally I cannot see the appeal of such vehicles, and of course there are other issues such as pedestrian safety, but on a purely environmental basis such restrictions on 4x4s do not seem justified. For example, the Land Rover Freelander diesel weighs 1600kg, is 4.4m long, and produces 205 g/km of CO₂ on the European drive cycle. A Mercedes S55 AMG weighs 2100kg, is 5.2m long and produces 317 g/km of CO₂. Now, the S55 AMG is an excellent and highly desirable car, but it would seem unfair to tax the Freelander more highly than the S55 on the basis that it is "less environmentally friendly", when by any objective measure it is not. It would seem particularly unfair to impose blanket restrictions/taxes on all 4x4s, without considering the technical features and fuel efficiency of their engines.)

Your main source of information in choosing an economical vehicle is the official fuel consumption figure, which is measured over a standard drive cycle. There are many different cycles depending on which country you live in, but the principle is the same. It is true that vehicles very rarely achieve the "official" figure in real-world driving, but in general if car A is better than car B on the official figures, it will also be better in the real world. (One personal observation is that very low-powered cars tend to be surprisingly poor in the real world, since they have to be driven aggresively to make good progress. Choosing a bigger engine may have less of a fuel economy penalty than the official figures would suggest.)

Given my scepticism about commonly available aftermarket fuel "saving" devices, and the genuine need to improve economy, I have been asked if it would be possible to retrofit any of the technologies described here to existing vehicles. The answer, unfortunately, is that it almost certainly would make no economic sense to do so.

Developing a more economical version of an existing engine by adding one of the technologies talked about above might typically cost £5 million. That sounds very large, but it needs to include:

• basic testing to prove the principle
• designing the new parts and the tools to make them
• producing a number of prototypes
• optimising the settings of the fuel saving device
• making corresponding adjustments to spark timing, fuel delivery, etc
• verifying that the vehicle still meets emissions standards
• adjusting and validating the On-Board Diagnostics on the vehicle
• testing the system in extremes of hot, cold and altitude
• durability testing
• etc

Now, suppose you plan to offer that technology as a retrofit kit for older vehicles instead. You'd still need to spend just as much on development, but probably spread over fewer vehicles, so that could be £100-£200 per vehicle. Critically, many major engine parts would probably need swapping (including most likely the engine management system (ECU)) and including labour you would be lucky to see change from £1000. That makes the total cost to the customer well over £1000 which, for the expected £50/year saving in fuel, makes no sense.

More logical is for buyers of new cars to choose ones fitted with advanced fuel-saving technology - these vehicles will slowly but surely replace the older, less economical ones, and so the fleet average economy will improve. Fortunately it is generally true that, while there are many very old cars still on the road, on average they are driven less than newer vehicles are. Replacing a small number of very heavily used vehicles with more economical ones can have a significant effect on overall fuel consumption.