Fuel trim patterns (advanced)

Fuel trims aren’t just “lean or rich.” They’re a pattern. If you compare idle vs load, bank 1 vs bank 2, and trims vs MAF, you can usually tell whether you’re chasing air, fuel delivery, or sensor scaling.

Before you read the numbers

Confirm closed loop and stable coolant temp (cold engines lie).
Know what trim you’re viewing: STFT reacts now, LTFT is the ECU’s learned correction.
Look at trims at idle and at a steady 2,500–3,000 rpm light load.

Rule of thumb

Use the sum: STFT + LTFT. A small LTFT with a big STFT can still mean a real issue — it just hasn’t learned yet.

Common trap

Trims can look “okay” because the ECU is masking the fault. Always pair trims with symptoms, misfires, O2 behaviour and load response.

Pattern table (use this like a cheat sheet)

Positive trims mostly at idle → unmetered air (vacuum leak), PCV stuck open, intake gasket leak.
Positive trims mostly under load → fuel delivery (weak pump, restricted filter), MAF under-reporting, exhaust restriction on some setups.
Negative trims at idle → leaking injector(s), high fuel pressure, purge valve stuck open (adding fuel vapour), MAF over-reporting.
Bank 1 and Bank 2 both similar → system-wide: MAF scaling, fuel pressure, purge, big air leak before split.
Only one bank off → bank-specific: intake leak on one side, injector imbalance, exhaust leak near one upstream sensor.

The 3 fastest “confirm” tests

Snap throttle test: STFT should swing rich then lean smoothly. Flat/slow response points to sensor or fuel delivery issues.
MAF sanity: compare g/s to engine size and rpm (and compare to known-good logs if possible). Big mismatch supports MAF scaling.
Purge test: clamp/command purge off temporarily. If trims drop back toward zero, you found a purge/vapour issue.

When trims scream “wiring/ground”

If multiple sensors read oddly (MAF, MAP, throttle, O2) and trims go wild with bumps/heat, don’t ignore shared references and grounds. Use: