Fuel Injector Flow Data: What 2,359 Bench Tests Show

May 8, 2026 · updated August 2026

By Ryan Bates, Marine Mechanic

Want your own set's before-and-after numbers? You can send them in here.

Most of what's written online about fuel injector cleaning is opinion. We keep our flow bench results in a format we can actually pull and add up, so here are the real numbers instead. Since 2022 we've logged 362 injector sets, 2,359 individual injectors, with before-and-after data on every one. This is that data, rough edges left in.

One honest note on the sample: 362 sets is our full digital record, August 2022 through August 2026. A handful of even older jobs predate our structured logging and aren't in here. Every number below comes straight off the bench, not from memory.

How We Test

Every injector goes on the same bench, a GB800 multi-port tester. On each one, before and after cleaning, we log four things:

• Resistance (ohms) - tells us if the solenoid coil is healthy
• 10-second open flow (ml) - injector held open, total flow capacity
• 60-second pulsed flow (ml) - injector pulsed at duty cycle, real-world fueling
• Leak test (pass/fail) - injector pressurized closed, checks the seal

Test pressure depends on the injector. Most of what we see runs at 40-50 psi, standard port-injection (MPFI) pressure. A handful are high-pressure GDI at 100-120 psi, and the occasional throttle-body (TBI) set around 15 psi. Pressure tells us the injector type, not what engine it came out of.

Balance Is the Whole Game

The number that actually matters is how evenly a set flows. For an engine to run smooth, every injector in the set needs to flow within a few percent of the others. When one is flowing 15-20% off from its neighbors, that's your misfire, your rough idle, your surge. We measure it as the spread between the highest and lowest injector in a set, as a percent of the set's average.

Across 346 sets with a complete before-and-after on every injector, here's how they showed up:

• 48% arrived more than 10% out of balance - bad enough to feel it
• 79% arrived more than 5% out - some noticeable variance
• The average set arrived with about 18% spread between best and worst injector

So a big chunk of what shows up has a real, measurable balance problem, not just a little dirt. Those engines are not going to run right on what they have.

About half the engines we see come in with injectors that don't match each other well enough to run smooth, and cleaning brings most of them back in line. That's the whole reason to do it.

Cleaning Cuts the Spread Roughly in Half

After cleaning, those same sets tightened up hard:

• Sets more than 10% out of balance: 48% down to 24%
• Sets more than 5% out: 79% down to 59%
• Average spread: about 18% down to about 9%

The ones that don't come all the way back usually have a worn pintle or seat that won't seal evenly, something no chemical or ultrasonic process can fix. The bench flags those so you know before you reinstall.

Why We Don't Brag About Flow Numbers

Here's where a lot of shops oversell. If you clean a set and just look at raw flow, it is not a clean "everything goes up" story. Across 2,211 injectors with matched before-and-after flow:

• 15% flowed noticeably higher (more than 5% up) - the badly clogged ones opening back up
• 21% flowed a little higher (1-5% up) - typical buildup coming off
• 25% didn't really change - already clean
• 39% actually flowed lower after cleaning

That last group throws people off, so let's be straight about it. An injector that flows lower after cleaning was usually flowing high because debris was holding the pintle open or gunk was skewing its calibration. Cleaning resets it to its true rate. If it was the high one in the set, it comes down to match the rest. That is exactly the point. The engine doesn't want high flow, it wants matched flow, which is why the balance chart above is the number to watch, not this one.

Cleaning is about getting your injectors to match each other, not making them spray more. Matched injectors are what let an engine idle smooth and pull clean.

Are Any Beyond Saving?

Sometimes, yeah. An injector with a shorted or open coil can't be cleaned back to life, and the bench catches it instantly on the resistance check. In this data, resistance ran from 1.0 to 45.0 ohms, and 8 injectors came back out of spec, almost certainly open or damaged coils. That's under half a percent of the 2,359, so dead injectors are rare, but they're real, and no amount of cleaning brings them back. When we hit one, it's flagged on your report before anything ships back, so you're not paying to clean a paperweight.

What Comes Through Our Shop

The bench data doesn't record what engine each set came out of, so this part is just from doing the work, not from the numbers: the bulk of what we clean is marine outboards. Day to day, the ones we see most are the shared CDH210 / INP-771 (Yamaha F75-F150, Suzuki DF90-140), the CDH275 / EAT152 (Yamaha F150 4-stroke), Keihin OX66 (Yamaha 250 V6), Mercury Optimax (fuel and air injectors), and Yamaha HPDI. The non-marine work is mostly Harley-Davidson and Mercruiser. We do very little general passenger-car work.

What This Means If You're Thinking About Sending Injectors In

If your engine is misfiring or running rough - there's a good chance you're in the group with a real balance problem. In our data, cleaning brought most out-of-spec sets back into line.

If it's preventive maintenance and nothing's wrong - expect modest changes. Worth it if you're already in there or repowering, less urgent if it's running fine.

Either way, you get the truth on paper. The report we ship back shows every injector's before-and-after flow, its leak result, and its resistance, so you're deciding on numbers instead of guesses. Here's what we look for in a healthy injector versus a bad one, and why a bottle of additive can't replicate what the bench shows you.