Who or what is American Racing Headers? They’re an American brand (you guessed right) that takes a lot of pride in making high performance, quality exhaust systems right here in the USA – 100% of it. They focus mostly on domestic vehicles, but offer a small line of products for foreign OEM’s like Lexus and BMW. Out of nowhere, they started producing special ARH E46 M3 headers which is 6 into 1 long tube header design for the S54. It consists of six individual 1.75″ primaries mating to one 3.5″ section 1 all made in T304 stainless steel with .049 wall. What does that do? The idea is that unlike the traditional 2-piece header, a 6 into 1 with their “patented rotational firing” design will optimize airflow (even more than the traditional headers) to produce significant power on the S54.

Oh yeah and also, sound! The design of the 6 into 1 should alter the sound ever so slightly, but in a way that it replicates the sound of a V8. We’ve all seen that mysterious video uploaded on YouTube in 2007 titled “Custom M3 Headers” right? I think it’s still a mystery to what exhaust setup it is, but apparently it is a 6 into 1 like the ARH.

Okay let’s move onto some numbers. American Racing Headers conducted their own tests comparing a stock E46 M3 S54, traditional header setup and ARH header setup. Both header setups include an intake and tune. They published the following results:

So when you put the data in a table, it looks something like this below. Delta (Δ) is comparing against stock run file.

I know going into this ARH setup, I knew I was going to most likely see a dip in torque – especially the low end of the RPM range. I knew this because I talked to 2 other owners and did a bunch of digging to set expectations before I pulled the trigger on the ARH. All the users I talked to and all the research I found led to strong signs of torque loss, but I still pulled through because…well I’ll tell you why exactly at the bottom of the read.

Onto my own results. I’m running the ARH headers, Bimmerworld ‘street/track’ section 2 mated to a Bimmerworld Race 3.5″ section 3. Supporting engine mods include a Turner Motorsport CSL Airbox and Kassel Performance MAP sensor all controlled by the MSS54HP ECU with a HTE custom dyno tune. Unfortunately no off-the-shelf tune will work with this setup and Hassan has a lot of experience tuning these headers so I chose him to tune this beast. Results below:

• Red line = baseline run (no tune) on 91 octane
• Blue line = tuned on 91 octane
• Green line = tuned on E30 ethanol blend (APE Flex Fuel kit)

So in terms of max power, it did great! It was actually on par with the ARH claim of 65 whp over stock as I made 64 whp over stock on 91 octane. I also made more than my previous Evosport ‘traditional’ header setup so all good right? Can’t answer that yet until we look a little further. Max horsepower is just a vanity number. What you really need to see is the dyno graph and how much of that power is sustained throughout the RPM range.

Here are the max numbers in a table view. I also added in my stock tune run file with only exhaust and intake. No headers. Delta (Δ) is comparing against stock run file.

This section of the article is now supportive of what the community claims have been: torque loss. I overlaid my previous setup with Evosport ‘traditional’ headers and E30 ethanol. All the others supporting mods are the same e.g. CSL airbox, Bimmerworld race exhaust, E30 ethanol, etc.

• Red line = American Racing headers
• Blue line = Evosport ‘traditional’ headers

Here are the max numbers in a table view. I also added in my stock tune run file with only exhaust and intake. No headers. Delta (Δ) is comparing against stock run file.

The loss in torque isn’t a major concern for me because this is a track-only vehicle where it lives in the 5k+ RPM range on track. This is my use case. For yours, it might be different. If you’re just mostly street driving then you’ll definitely feel a difference in power down low if you’re coming from a traditional style header setup.

So back to my track use-case. According to the dyno results, I’ll see significant torque loss if I am below 5k RPM and even if I do dip down I don’t believe it will have significant impact on my lap times. For the areas of the track where I am dipping down to the low RPM range, I can also try shifting one gear higher to keep the car in higher power range. Also worth to note, I am running shorter 3.91 gears so naturally the car wants to be in a higher RPM range which helps.

I just recently took this ARH E46 M3 headers setup to the track to see real world results and the data shows I am mostly in the 5k+ RPM range. The car pulls hard up top and I do think it’s faster than the previous traditional setup. Sacrifice low end torque, that I’m rarely in, for high end horsepower. Even when driving mindlessly around the track there wasn’t a section where I felt down on torque. Actually didn’t come across my mind at all on track even during the 2 slowest parts of the track where I leave it in a lower gear normally. This thing pulls hard all the way to redline.

At first when I saw the dyno results I had second thoughts about sending it back to EAS and telling them to put my old headers back on. But I had to give it a shot on track. Yes, the torque just disappears down low. Yes, even with a tune the low end torque is unrecoverable. However, after actually driving it on track I am very happy and will not go back. This ARH is here to stay.

Lastly, another reason why I am keeping it. The sound! I think it sounds so much better. I mean, part of modifying is not only just to go faster but a balance of form and function. You’re not going to get this type of sound with any other traditional header so that alone makes a car so much more unique. Take a listen for yourself below and if you have these headers, please share a link down below in the comments so I can hear other setups!