I have 3 tire pressure gauges that give readings up to 2 psi different in the 30-40 psi range: a pencil-style gauge, an electronic gauge, and a dial gauge.

How does one determine which is more accurate? Is there a way to check them against a reference?

My guess is that any other gauge I have easy access to for comparison (filling station, tire store, home air compressor), won't serve as a good reference.

Due to the variance between the gauges, I've settled on always using the same one (the dial gauge, as it seems most repeatable). Using it I can pretty well tell the variation in pressure between the 4 tires. But, I don't know how much high or low the gauge reads.

I found the following threads about gauges, but I didn't find any ideas about testing gauges.

https://www.northamericanmotoring.com/forums/showthread.php?t=31219

https://www.northamericanmotoring.com/forums/showthread.php?t=9797

Any ideas?

I just use the pencil guage I bought about 15 years ago before those
became the $3 pieces of crap. It's within 1psi vs the other guages I have
and used, so I use it as my main reference guage. Since I go for the
measurement rather than the actual PSI accuracy, it serves my purpose.
I set my tire pressures based on experience of that particular reading
knowing that it is no more than 1psi off the other guages.

I did not like those pensil gauges at all. I've used an electric one for the past few years and have had no problems. That's probably the most accurate one. I just recently purchased a 0-60 psi dial gauge and have found it very accurate compared to the electric one. I like the dial one best because it has a seperate bleeder valve.

Generally you get what you pay for.

A good sturdy dial faced tire pressure gauge with a bleeder valve is a good place to start. I have a basic one by Longacre bought at ebay that works well. When comparing it to other gauges they can be off up to 6 psi so it's good to compare with other sources.

Even if your gauge is off by 1 or 2 psi you can record what might work for you and estimate how much it really might be. It doesn't really matter for street use but for track and autocross it might make a difference. Using the same gauge is key as well as recording the results. Every tire has a different pressure for best performance.

Setting a good psi for your tires on the track can be done based on tire temperatures using a temp gauge and chalking of tires.

Bought a dial one with a entended hose and a bleed off valve for 16.00 from Pep Boys. It works awesome. I first saw these while at the track when the professionals were racing. Almost bought the same thing from NAPA but for 65.00!! Whoops! That was close. Don't be tempted just to throw money at something for the sake of having the most expensive one. The first one I bought from Pep boys was only 11.00, but I quickly returned it (tested it in the parking lot) when it leaked air where the adapter fitted to the valve stem. I went back in and got their most expensive non-digital one on the shelf and it was only 16.00. I can read it down to the 1/2 pound. Gets very repeatable results everytime!

that no matter what you buy or how much you spend, any instrument is subject to wear and calibration drift. Any instrument must be calibrated periodicaly if one is to trust comparisons over time.

At the company I work for, we make (among many other things) pressure guages that are very, very accurate and very, very expensive. Most guages are calibrated against reference guages that are calibrated periodically agains NIST traceable standards, or first principles measurement standards (whick are really rare, but wonderful when found, like atomic clock time standards).

I've found it's easier to get a good dail guage than a good electonic guage. One a dial, you can infer about 10% of the smallest measurement, and the cheeper electrical are only good for 0.5 psi or so.

So no matter what you bought, you should check it every so often.

Matt