How Driving Style Affects Fuel Consumption: What the Research Actually Shows

Put two drivers in the same car, on the same route, in the same traffic — and their fuel bills can differ by a third. That is not folklore: it is one of the most consistently replicated findings in vehicle-efficiency research. The US Department of Energy puts the cost of aggressive driving at 15–30% at highway speeds and 10–40% in stop-and-go traffic; eco-driving training programs across Europe reliably squeeze 5–15% out of ordinary commuters without touching the car. This article walks through what the studies actually measured, why the physics works that way, and which habits matter most — ranked by the size of the number, not by how often they appear in listicles.

The headline numbers

Three sources anchor everything below. First, instrumented-vehicle work at Oak Ridge National Laboratory, which drove identical cars through identical cycles with calm and aggressive throttle profiles and found the aggressive style burning up to a third more fuel on the highway and up to 40% more in city traffic. Second, the eco-driving training literature: dozens of before/after studies of ordinary drivers show 5–15% savings immediately after a course, decaying to a still-real 3–7% months later unless the driver gets ongoing feedback — a decay curve that says habits, not knowledge, are the hard part. Third, national fleet programs like the Dutch "Het Nieuwe Rijden", which moved an entire country's average consumption by roughly 3% — small per driver, enormous at scale. The spread between those numbers is the honest answer: how much style matters depends on how bad your baseline is. A patient driver has little to gain; a hurried one is carrying a 20–30% surcharge every day.

Where the fuel actually goes

Every result below follows from three pieces of physics. Braking throws energy away: all the fuel spent accelerating your car to speed is converted to heat the moment you touch the brake pedal — in city driving, roughly half of all fuel goes into accelerations, so every avoidable stop is fuel spent twice. Aerodynamic drag grows with the square of speed: pushing air aside at 130 km/h (80 mph) takes about 70% more force than at 100 km/h (62 mph), which is why speed dominates highway consumption. An idling engine divides by zero: fuel burned while the car isn't moving makes your per-distance consumption arbitrarily worse — half a litre to a litre (a quarter of a gallon) per hour, doing nothing. Keep those three in mind and every number below stops being a trivia fact and becomes predictable.

Habit 1: highway speed — the biggest single lever

Fuel economy for most cars peaks between 70 and 90 km/h (45–55 mph) — fast enough to be in top gear at low rpm, slow enough that drag hasn't taken over. Above that, the square law bites: each additional 10 km/h (6 mph) above 100 km/h costs roughly 7–10%, and holding 130 km/h instead of 110 typically costs 15–25% on the same stretch of road. The US DOE translates it into money: every 5 mph above 50 mph is like paying about a quarter more per gallon. The research also measures what the extra speed buys you: on a 300 km (190 mi) leg, 130 km/h instead of 110 saves about 25 minutes of driving — before the extra fuel stop.

Habit 2: how you brake matters more than how you accelerate

The instinct is to blame hard acceleration, but the energy accounting points at the other pedal: acceleration only wastes fuel when the speed it bought is thrown away at the next brake application. That is why anticipation is the single most effective city-driving skill in every eco-driving curriculum: look 10–15 seconds ahead, lift off early when the light is red or traffic is bunching, and let the car coast to the obstacle in gear. On any modern injected engine, coasting in gear activates deceleration fuel cut-off — the injectors shut entirely and consumption reads 0.0 — while coasting in neutral burns idle fuel the whole way down. Drivers who learn only this one habit — trade late braking for early lifting — show 10–15% city savings in follow-up measurements, and their brake pads quietly last twice as long.

Habit 3: idling — half a litre an hour for standing still

A warm idling engine burns roughly 0.5–1 litre per hour (a quarter to half a gallon), and up to double that with the air conditioning working. Measurements at Argonne National Laboratory settled the restart question years ago: beyond about ten seconds of standing, shutting off and restarting uses less fuel than idling through — a modern starter and injection system make the restart cost trivial. That is exactly the arithmetic behind factory stop-start systems, which recover 5–8% in city cycles. The practical rule: railway crossings, waiting at the school gate, drive-through queues — off. Ordinary traffic lights — let the stop-start system do its job.

Habit 4: cold starts and short trips — the hidden multiplier

A cold engine runs deliberately rich and fights thick oil, so the first kilometres cost up to twice the warm consumption — and a trip short enough that the engine never fully warms up is driven entirely inside that penalty. US EPA testing puts a conventional car's economy about 15% lower at −7°C (20°F) than at 25°C (77°F), and up to 24% worse on trips of 5–6 km (3–4 miles). Driving style here means trip planning: chain errands into one loop instead of three cold starts, do the longest leg first so everything after runs on a warm engine, and skip the driveway warm-up — thirty seconds and gentle driving warms a modern engine faster and cheaper than ten minutes of idling (which is Habit 3 wearing a winter coat).

Habit 5: steady throttle — and what cruise control gets right

Instrumented studies of throttle traces show that "speed-wobbling" — unconsciously surging and lifting between 95 and 115 km/h — costs measurably more than holding a steady 105, because every surge is a small acceleration paid for at highway drag. On flat roads cruise control beats a human foot precisely because it is boring. The footnote from the same studies: on rolling hills, dumb cruise control gets it wrong — it downshifts and charges uphill to defend the set speed, where an attentive driver would let a few km/h bleed off on the climb and recover them for free on the descent. Flat motorway: cruise on. Hilly road: your right foot, driven by the same anticipation as Habit 2.

Habit 6: the acceleration myth — crawling isn't the answer

The classic eco-tip says "accelerate as slowly as possible", and the engine-efficiency data says otherwise. Combustion engines are most efficient under moderate-to-brisk load at low-to-middle rpm — the worst place to operate one is barely-open throttle, which is exactly where glacial acceleration lives. The research-backed technique: accelerate purposefully — roughly half to two-thirds throttle — up to your cruising speed, shifting (or letting the automatic shift) early, then hold steady. You spend less total time in the inefficient transition and arrive at the efficient cruise sooner. What the data does condemn is flooring it: wide-open throttle commands a rich mixture that dumps extra fuel for power, and on many engines briefly stops obeying efficiency logic entirely. Brisk, not brutal — and never to the next red light, which hands everything straight back to Habit 2.

Habit 7: air conditioning and open windows

The compressor is a real mechanical load: 3–10% in typical conditions, and 20% or more in extreme heat on short city trips where it works hardest against a sun-soaked cabin. Open windows cost drag instead — negligible in town, but climbing with the square of speed like everything aerodynamic. The crossover the wind-tunnel and track studies agree on: below about 70–80 km/h (45–50 mph) open windows are cheaper; above it, A/C wins. The free trick that beats both: ventilate the parked sauna for a minute before driving off, so the compressor cools air, not the dashboard.

What your style costs in money

Take an ordinary case: 15,000 km (9,300 mi) a year, a car that averages 8 l/100 km (29 mpg) driven calmly. A consistently aggressive style at the modest end of the ORNL range — 20% — burns an extra 240 litres (63 gallons) a year: at typical 2026 fuel prices, roughly $220–260 in the US and €400–450 in Western Europe, every year, for arriving a few minutes earlier and replacing brake pads twice as often. Run your own numbers with your own consumption — and if you don't know your real figure, that is the first fix: the trip computer flatters by 5–10%, while fill-up-to-fill-up arithmetic doesn't lie.

Three myths the data keeps killing

• "Coasting in neutral saves fuel." Backwards on any injected car: in gear with your foot off the pedal, fuel cut-off means literally zero consumption; in neutral the engine burns idle fuel to keep itself spinning — and you've given up engine braking and instant response. It survives from the carburetor era.
• "The engine must warm up before driving." Ten minutes of driveway idling is the most expensive way to warm an engine — slow, rich-mixture burn at zero distance. Thirty seconds to circulate oil, then gentle driving: warmer sooner, litres cheaper. (Defrosting windows is a visibility job, not an engine one.)
• "Premium fuel cuts consumption." If your car is designed for regular, higher octane changes essentially nothing — octane is knock resistance, not energy content. The full breakdown, including the cars where octane does matter, is in our article on regular vs premium fuel.

Measure first, then change one thing

Every eco-driving study that achieved lasting savings had one design feature in common: the driver could see their own number. Feedback, not lectures, is what keeps the 10% from evaporating. So run the experiment on yourself: log your real consumption between full fill-ups for two or three tanks, pick one habit from this list, drive two more tanks, compare. Brimly does the bookkeeping for you — scan the receipt, and it computes true consumption between full tanks, shows the factory figure alongside, and draws the trend so the effect of a changed habit is visible within a couple of weeks.

And if the number refuses to move no matter how gently you drive, the cause may be the car, not the driver — our free car directory lists the documented consumption figures and known weak points for 200+ models, and the free AI assistant can tell you whether your figure is normal for your exact engine. More hands-on fixes are in the companion piece: 12 proven ways to reduce consumption in an automatic car.