ARAI (Automotive Research Association of India) mileage is the fuel-economy figure tested in a controlled lab environment under a standardised driving cycle. Real-world mileage — what you actually get on your commute — is typically 20% to 35% lower, because the ARAI test uses ideal conditions, no air-conditioning, fixed speeds, and a light load. Trust the ARAI number for comparing cars against each other; never trust it as a budget figure for your own fuel bill.
| Segment | Typical ARAI claim | Typical real-world | Shortfall |
|---|---|---|---|
| Small petrol hatchback | 22-24 km/L | 15-17 km/L | ~30% |
| Mid-size petrol sedan | 18-20 km/L | 12-14 km/L | ~30% |
| Mid-size petrol SUV | 16-18 km/L | 10-12 km/L | ~35% |
| Diesel sedan | 22-25 km/L | 17-19 km/L | ~22% |
| CNG hatchback | 30-32 km/kg | 22-25 km/kg | ~25% |
| EV (ARAI range) | 400-500 km | 280-360 km | ~28% |
What ARAI actually tests
ARAI runs every new car sold in India through a standardised chassis-dynamometer test — a treadmill-like rig that lets them measure fuel consumption with the car stationary. The protocol is the Modified Indian Driving Cycle (MIDC), which simulates a ~20-minute drive cycle covering both urban stop-start and extra-urban steady-speed phases.
Important features of the test:
- Ambient temperature is controlled (typically 25-30°C — no peak Indian summer, no winter start).
- Air-conditioning is OFF throughout.
- Only the driver, no passengers, no luggage.
- Accessories like headlights, music, wipers are OFF.
- Tyre pressure is set to manufacturer spec; zero wear.
- Acceleration and deceleration follow a prescribed profile — not aggressive, not slow, just prescribed.
- Top speed is capped at 90 km/h.
The output is a single fuel-economy figure (km/L, km/kg, or km/kWh) that gets printed on the brochure and the car's window sticker. It's comparable — every car is tested on the same cycle — but it's not representative of any real driver's usage.
Why your real mileage is lower
- Air-conditioning (the biggest single factor in India): costs ~1-3 km/L depending on climate. Running full AC through summer can knock 15-20% off the ARAI figure alone.
- Traffic stop-start: ARAI's urban phase has prescribed idling. Real Indian city traffic has far more idling, creeping, and short acceleration bursts — all fuel-hungry.
- Highway speeds above 90 km/h: aerodynamic drag rises with the square of speed. Sustained 110 km/h cuts mileage 8-12% vs 80 km/h. ARAI never tests beyond 90.
- Passenger and luggage load: every 50kg of extra load costs ~1-2% mileage.
- Tyre pressure below spec: under-inflated tyres raise rolling resistance; 5 psi under-inflated costs 2-4% mileage.
- Rooftop accessories (roof rails, boxes): worst case can cost 5-8% on highway runs.
- Cold starts and short trips: engines run rich until they hit operating temperature. A cluster of sub-5-km trips can halve your effective mileage.
- Driver style: aggressive acceleration and hard braking cost 10-25%. ARAI's cycle is deliberately moderate.
Stack these effects and you see why the 20-35% shortfall is typical, not an aberration.
ARAI mileage is still useful — when you compare
The ARAI figure's purpose isn't to predict your fuel bill — it's to give buyers a consistent yardstick to compare two cars. If Car A is rated 22 km/L ARAI and Car B is rated 18 km/L, Car A will likely give you ~20% better real-world mileage too, even though neither will hit the ARAI number.
The ratio holds; the absolute number doesn't. Treat it like you'd treat a smartphone's "up to 48 hours battery life" claim: useful for ranking, useless as a schedule.
What number to use for cost calculations
For TCO or EV-vs-ICE break-even math, use the realistic number for your usage pattern — not ARAI. Rule of thumb:
- Highway-heavy driver (> 60% of km on expressways, moderate speeds): apply a 15% discount to the ARAI figure.
- Mixed metro + highway: 25% discount.
- Pure urban commute (bumper-to-bumper, AC on): 30-35% discount.
- EV range (ARAI): apply 25-30% discount for highway ranges, 15-20% for steady-speed drives.
Our calculators default to these realistic numbers, not ARAI claims — so the TCO output is what you'd actually pay.
Why CNG and diesel gaps are narrower
Diesel engines produce more torque at low RPM, so they handle stop-start urban traffic more efficiently than equivalent petrols — the real-world shortfall tends to be ~22% rather than ~30%. CNG similarly holds up well because it's often driven in low-load urban conditions that match its combustion profile.
EVs have the largest relative shortfall in range terms, partly because their ARAI test doesn't include climate control or realistic highway speeds — both disproportionately affect battery draw.
How to measure your actual mileage
- Fill the tank completely. Note the odometer.
- Drive normally for 400-600 km.
- Refill to the same fill-point at the same pump (same nozzle click-off level matters).
- Divide km driven by litres refilled.
Do this across 3-4 tanks to smooth out variance. Most owners find their real mileage converges to within ±1 km/L of the average over a few tanks. The on-board MID display is usually 5-10% optimistic vs the tank-fill method — trust the manual calculation.
People also ask
Is ARAI mileage accurate in city driving?
No — ARAI's urban phase runs a prescribed, moderate-pace cycle with no AC load, not real Indian metro traffic. City real mileage is typically 25-35% below ARAI. Highway real mileage is closer — usually 10-20% below ARAI at moderate cruise speeds.
Why do manufacturers quote ARAI numbers if they're unrealistic?
Because it's the legally mandated comparison standard. Every new car must display its ARAI-certified figure; it's what allows regulators and customers to compare cars on equal terms. Real-world mileage varies by driver, so there is no "true" number to publish.
Does running AC always reduce mileage?
Yes — AC compressor draws engine power. Average cost is ~1-3 km/L on petrol/diesel and ~10-15% of range on EVs. Well-maintained AC systems cost less than neglected ones; the compressor clutch engaging repeatedly on a low-refrigerant system is worse than continuous operation.
Which car has the highest real-world mileage in India?
On current (2026) benchmarks: mild-hybrid small petrols (Maruti Grand Vitara hybrid variants) and strong-hybrid Toyotas deliver the highest real-world figures among ICE cars — often 22-25 km/L city. EVs effectively beat everything on ₹-per-km but range-per-charge is a different metric from mileage.
Do EVs have the same ARAI-vs-real gap?
Yes — and often a larger one in percentage terms. ARAI EV range figures commonly overstate real-world range by 25-30%, particularly on highway drives with HVAC on. Use the realistic range (~70% of ARAI) when calculating EV total cost of ownership.