Why you should never trust the test/reset buttons on a GFCI

Rather eye opening :shock:

How its supposed to be:

I didn’t go to your links because (based on the title of the thread) I have experience with defective GFCI devices.

I will repeat what I’ve been saying for years - you must verify that the circuit has de-energized when testing GFCI devices. The “snapping” of the test button is not a definitive test to determine that the GFCI device is functioning properly.

I’ll quote Russell Ray on this one, “I test GFCI outlets using the manufacturer’s test button, just like the manufacturer recommends in their installation instructions. If it’s good enough for the manufacturer, it’s good enough for me.”

That is great, but do you have a tester in the receptacle to make sure the power went out?

True. Just last week, a receptacle in a bathroom did not respond to my test button on my little checker. As always, whether the test button works or not, I manually “tripped”, it, and the reset popped out, but the receptacle remained energized.

This is what I am saying, the manufacturer’s recommended test means nothing.

If you push the test button and it is still energized then the receptacle is defective. End of story. Write it up and carry on. I do not see why there is controversy.

Not everyone checks to make sure its truly dead. The only way to verify that is via plug in tester, one that checks both the line to neutral and line to ground. Granted most inspectors use such a device, however its a good idea to know that it certainly is possible to have a failure mod where the GFCI trips, resets and indicates the correct lights whilst being live the whole time.

Maybe this is the first time someone has seen something mechanical fail?

Ok, that would make sense. I am not sure why you would not check it with a tester once tripped. That seems ridiculous to me.

Greetings Everyone,

I have also seen my fair share of defective products over the years. It’s just par for the course (as sad as that shall seem). The fact that GFCI’s have saved (documented) thousands if not hundreds of thousands of lives should be indication of their value in of itself.

UL 943 has changed over the years and that group have made many strives in advancement but anything or everything could and will fail at some point so I don’t see much controversy here.

The issue is how detailed is an HI going to go?..If an HI did these types of tests and reported that I would have a field day with them in court the time something happens because I would argue they affected it’s ability to function properly after the fact…but alas thats the “Win at all costs Paul” and not the " It’s ok…no worries Paul" approach.

I love the videos…that Paul LOVES It…but I caution HI’s from going that far in how they report stuff. The manufactures state that these devices are the be tested on a monthly basis, when they are not (which is not documented) they have a way OUT in court…just sayin…been there…done that.

The best defense for the HI is to test per manufacturers instructions and report your findings and move on. The moment you become overly involved is the moment you begin to hear court room bells ringing in my opinion.

But again…being an analytical guy…I did love the videos.

So HI should just ignore this failure mode in fear of lawsuit?

Well it probably won’t help when you have said that you will ignore the manufacturer instructions.

I believe that most manufactures require the device to be tested under a load. The handheld tester would qualify as a load, and will alert if power was indeed turned off.


Where did I say that? All I said indirectly was that they are simply not enough and it has been proven several times.

Such does not test line to ground where as a plug in tester does.

A plug in tester is worthless on an ungrounded circuit. They are also not the recognized testing method, regardless of what you think.


Most homes built in the last 40 years have fully functional EGC through out. We all know that.

I am well aware of what manufacturers recognize as a testing method; otherwise I might not be starting this thread in the first place. However reality says otherwise, and remember that reality is not what I imagine or others imagine.

So the notion I am gaining is that there would be nothing wrong with inspectors missing this type of failure mode? Again, I never said there instructions be ignored, simply they be double checked with a plug in tester to verify all lights have gone out.