PullScan
Fueling

What Causes High Fuel Trims?

High fuel trims show that the ECU is adding fuel, but they do not identify the cause by themselves. Learn how air leaks, fuel delivery, ethanol content, sensor inputs, and supplemental fueling affect trim behavior.

By PullScan Engineering · Published July 26, 2026 · Updated July 26, 2026

High fuel trims mean the engine controller is adding fuel in response to feedback indicating that the mixture is leaner than the controller intended.

That is an important distinction: fuel trim is a correction, not a diagnosis.

A high positive trim does not automatically prove that the engine has a failing fuel pump, an intake leak, or a bad oxygen sensor. It shows what the controller is doing. Finding the cause requires examining the operating window and the related signals.

What does fuel trim measure?

The engine controller first estimates how much fuel is required from inputs such as airflow, throttle position, temperature, and engine speed. Once oxygen-sensor feedback is available, the controller adjusts fuel delivery to move the measured mixture toward its commanded value.

Positive correction means the controller is adding fuel. Negative correction means it is removing fuel.

Kia’s technical training material describes short-term fuel trim as the controller’s rapid correction based on oxygen-sensor feedback. Long-term adaptation changes more slowly and helps return short-term correction toward the center of its available range.

JB4 logs require an additional platform-specific detail. On the Kia/Hyundai JB4 scale:

  • 25 represents approximately zero correction
  • 50 represents approximately +34% correction
  • 0 represents approximately −34% correction

Therefore, a JB4 trim value of 40 must not be described as 40 percent.

What can cause high positive fuel trims?

High positive trims generally mean the controller believes additional fuel is required. Possible categories include:

Unmetered or incorrectly measured air

An intake leak can allow air into the engine that was not represented correctly in the controller’s original fuel calculation.

An exhaust leak upstream of an oxygen sensor can also affect the sensor’s interpretation of the exhaust stream.

Insufficient fuel delivery

Fuel pressure or fuel-flow limitations can prevent the commanded quantity from reaching the cylinder. Possible areas include pumps, restrictions, injectors, wiring, and fuel-pressure control.

The trim value alone cannot identify which component is responsible.

Incorrect sensor information

Airflow, oxygen-sensor, temperature, pressure, or other input errors can cause the controller’s calculated fuel requirement to differ from the measured result.

A mixture-adaptation fault does not automatically mean that the oxygen sensor itself has failed.

A change in fuel composition

Ethanol has different chemical and energy properties from gasoline. The U.S. Department of Energy lists a stoichiometric air/fuel ratio of approximately 9:1 for ethanol and 14.7:1 for gasoline.

As ethanol content increases, a compatible calibration must supply the appropriate additional fuel. Otherwise, positive correction may rise.

Supplemental fueling that is not correctly balanced

A charge-pipe injection system adds another source of fuel. Burger Motorsports describes its Stinger CPI system as a seventh injector whose flow is controlled through the JB4 integration.

If supplemental fuel delivery is reduced too far, the factory controller may respond with higher positive trims. If supplemental fuel delivery is excessive, trims may move below their neutral value as the controller removes fuel.

This relationship is visible in the following real tuning sequence.

Real case: Kia Stinger GT 3.3T with CPI

The vehicle was running a JB4-controlled CPI kit. The reviewer changed only the requested fuel-bias setting between logging sessions and reviewed the two fuel-trim banks while the JB4 Meth PID reported CPI activity.

Initial setting: fuel bias 80

With CPI flow reported at 80, both trim channels moved well below the JB4 neutral value. The selected interval averaged 10.9 for bank 1 and 11.7 for bank 2.

The reviewer interpreted this as excessive supplemental fueling and requested a lower fuel-bias setting.

Fuel Bias 80: Trims Moved Below the JB4 Neutral Value

10 selected points
Line graph showing CPI flow fixed at 80 while both JB4 fuel-trim channels fall from approximately 20 to approximately 8.Elapsed time (s)JB4 scaled value%-1.0514.23
Fuel trim — bank 1 (JB4 scaled value)Fuel trim — bank 2 (JB4 scaled value)CPI flow — Meth PID (%)
With CPI flow reported at 80, the two JB4 fuel-trim channels fell from 20 and 23 to 8 and 8. The case reviewer interpreted this response as excessive supplemental fueling and recommended reducing fuel bias.

Limitation: This graph shows ECU correction behavior during one CPI-active interval. It does not independently prove injector flow, commanded lambda, or a mechanical fault.

View representative measurements
Elapsed timeFuel trim — bank 1Fuel trim — bank 2CPI flow — Meth PID
0 s20 JB4 scaled value23 JB4 scaled value80 %
2.35 s16 JB4 scaled value19 JB4 scaled value80 %
4.14 s13 JB4 scaled value13 JB4 scaled value80 %
5.06 s10 JB4 scaled value13 JB4 scaled value80 %
5.51 s9 JB4 scaled value10 JB4 scaled value80 %
7.74 s9 JB4 scaled value7 JB4 scaled value80 %
10 s8 JB4 scaled value8 JB4 scaled value80 %
10.52 s8 JB4 scaled value8 JB4 scaled value80 %
12 s8 JB4 scaled value8 JB4 scaled value80 %
13.18 s8 JB4 scaled value8 JB4 scaled value80 %

Sanitized CSV records 40–49 · parser 1.0.0

Fuel bias 40

After fuel bias was reduced to 40, the response moved in the opposite direction.

During the selected CPI-active interval, bank 1 averaged 41.4 and bank 2 averaged 41.6. Both were clearly above the JB4 neutral value of 25, meaning the controller was adding fuel.

The reviewer described the change as an improvement from the original overfueling condition, but considered the adjustment too large.

Fuel Bias 40: High Positive Trim Response

8 selected points
Line graph showing CPI flow rising toward 48 while both JB4 fuel-trim channels remain in the upper 30s and low 40s.Time (s)JB4 scaled value%-0.749.96
Fuel trim — bank 1 (JB4 scaled value)Fuel trim — bank 2 (JB4 scaled value)CPI flow — Meth PID (%)
After fuel bias was reduced to 40, CPI flow rose from 3 to 48 while both JB4 trim channels remained between 37 and 45. The elevated correction showed the ECU adding fuel relative to the JB4 neutral value of 25.

Limitation: The elevated trims identify ECU compensation, not the underlying cause by themselves. This comparison was not conducted under laboratory-controlled conditions.

View representative measurements
TimeFuel trim — bank 1Fuel trim — bank 2CPI flow — Meth PID
0 s37 JB4 scaled value39 JB4 scaled value5 %
0.99 s41 JB4 scaled value39 JB4 scaled value3 %
1.99 s41 JB4 scaled value42 JB4 scaled value16 %
3.61 s44 JB4 scaled value45 JB4 scaled value44 %
5.74 s42 JB4 scaled value42 JB4 scaled value39 %
6.75 s42 JB4 scaled value42 JB4 scaled value48 %
8.56 s42 JB4 scaled value42 JB4 scaled value46 %
9.22 s42 JB4 scaled value42 JB4 scaled value36 %

Sanitized CSV records 27–34 · parser 1.0.0

Fuel bias 45

Fuel bias was then increased to 45.

In the selected interval, both trim channels averaged approximately 30.5 and remained close to one another. This was a more moderate correction than either the low values seen at bias 80 or the elevated values seen at bias 40.

The reviewer accepted this result for the next logging step.

Fuel Bias 45: A More Moderate Trim Response

14 selected points
Line graph showing CPI flow increasing while both JB4 trim channels remain close together and generally nearer the neutral value of 25 than in the fuel-bias-40 test.Elapsed time (s)JB4 scaled value%-1.5520.98
Fuel trim — bank 1 (JB4 scaled value)Fuel trim — bank 2 (JB4 scaled value)CPI flow — Meth PID (%)
With fuel bias increased to 45, CPI flow ranged from 20 to 50 while the two trim channels averaged approximately 30.5 and tracked closely. The case reviewer accepted this intermediate setting for the next test.

Limitation: Closer-to-neutral trims do not establish that the calibration is safe or optimal. AFR targets, fuel composition, adaptations, and other operating conditions remain relevant.

View representative measurements
Elapsed timeFuel trim — bank 1Fuel trim — bank 2CPI flow — Meth PID
0 s34 JB4 scaled value34 JB4 scaled value31 %
1.93 s36 JB4 scaled value37 JB4 scaled value50 %
2.93 s38 JB4 scaled value37 JB4 scaled value50 %
5.44 s36 JB4 scaled value37 JB4 scaled value50 %
6.97 s35 JB4 scaled value34 JB4 scaled value50 %
8.05 s35 JB4 scaled value34 JB4 scaled value48 %
10.02 s30 JB4 scaled value31 JB4 scaled value36 %
11.93 s26 JB4 scaled value29 JB4 scaled value27 %
12.93 s26 JB4 scaled value25 JB4 scaled value20 %
17.18 s23 JB4 scaled value24 JB4 scaled value38 %
17.94 s22 JB4 scaled value23 JB4 scaled value50 %
19.43 s22 JB4 scaled value22 JB4 scaled value42 %

Sanitized CSV records 27–40 · parser 1.0.0

Map 4 verification

The same fuel-bias setting was then tested on Map 4.

The two banks continued to track closely in the selected CPI-active interval. The reviewer accepted the fueling response while separately noting a timing event later in the run.

Map 4 Verification with Fuel Bias 45

7 selected points
Line graph showing CPI flow between approximately 33 and 50 with the two JB4 fuel-trim channels tracking closely in the low-to-upper 30s.Elapsed time (s)JB4 scaled value%-0.8411.34
Fuel trim — bank 1 (JB4 scaled value)Fuel trim — bank 2 (JB4 scaled value)CPI flow — Meth PID (%)
On Map 4 with fuel bias still at 45, both banks continued to track closely during the selected CPI-active interval. The case reviewer accepted the log while separately noting a timing event later in the run.

Limitation: This interval documents one accepted review step. It is not a universal target range or a recommendation to use Map 4 or fuel bias 45 on another vehicle.

View representative measurements
Elapsed timeFuel trim — bank 1Fuel trim — bank 2CPI flow — Meth PID
0 s33 JB4 scaled value33 JB4 scaled value33 %
3 s36 JB4 scaled value35 JB4 scaled value50 %
4 s36 JB4 scaled value37 JB4 scaled value50 %
5.55 s38 JB4 scaled value38 JB4 scaled value50 %
7.5 s34 JB4 scaled value36 JB4 scaled value50 %
8.5 s34 JB4 scaled value36 JB4 scaled value50 %
10.5 s31 JB4 scaled value34 JB4 scaled value50 %

Sanitized CSV records 25–31 · parser 1.0.0

What this case demonstrates

This case does not establish that fuel bias 45 is correct for every Kia Stinger or every CPI installation.

It demonstrates a diagnostic method:

  1. Identify the interval where the supplemental system is active.
  2. Examine both trim banks during that interval.
  3. Change one control variable at a time.
  4. Repeat the log under comparable conditions.
  5. Look for consistent movement in the expected direction.
  6. Validate AFR or lambda, fuel pressure, timing, and other relevant signals separately.

The comparison also shows why “high fuel trims” cannot be interpreted in isolation. In this case, the higher trims at fuel bias 40 were consistent with insufficient supplemental fuel relative to the other tested settings. In another vehicle, high trims could result from an air leak, fuel-delivery limitation, sensor problem, or fuel-composition change.

High trims do not necessarily mean measured AFR is currently lean

A controller can add enough fuel to keep measured AFR near its intended value while fuel trims remain positive.

In that situation, the trims reveal the amount of correction being requested. They do not necessarily mean the final measured mixture is still lean.

To calculate a target-versus-actual fueling error, a log needs a documented commanded AFR or lambda channel with compatible units. These case files contain measured AFR channels but do not contain a validated commanded fueling target. A target error should therefore not be invented.

What about timing corrections?

Timing corrections must be evaluated separately.

Burger Motorsports documents the Kia/Hyundai Ign_2 through Ign_6 channels as cylinder timing reductions on compatible firmware when configured appropriately. It also notes that occasional corrections can occur normally and that repeated correction in the same cylinder within one gear is more meaningful than an isolated value or a shift-related event.

The fuel-trim sequence does not prove that fueling caused the reported timing corrections. Establishing that relationship would require a separate, synchronized analysis of timing, load, boost, fuel quality, temperature, and the exact correction window.

Practical investigation order

When high trims appear:

  1. Confirm the logger’s channel definition and scaling.
  2. Identify the relevant load, RPM, throttle, and supplemental-fueling interval.
  3. Compare both banks.
  4. Compare actual lambda or AFR with a documented commanded value when available.
  5. Check fuel pressure and delivery behavior.
  6. Inspect for intake or upstream exhaust leaks.
  7. Validate airflow and oxygen-sensor inputs.
  8. Confirm fuel composition and ethanol content.
  9. Make one controlled change and repeat the measurement.

Do not replace parts or increase boost based on a trim number alone.

Sources