I'm installing GFCI plugs in my house where some two-prong outlets used to be.

The outlet (leviton) reads as "green" (which the manufacturer says indicates successful installation) but I'm confused on why the gfci doesn't actually trip when I use a testing tool.

When I plug in the tool, the initial read I get is "open ground". That makes sense considering there's no equipment ground. What's strange though is that when I hit the gfci test button, the gfci trigger button doesn't trip and instead, the tool I'm using reads "hot/neutral reverse" (only when I'm pressing the gfci button).

Any ideas on what's going on?

tester with gfci button held

  • 4
    Those "magic 8 ball" testers are laughably unreliable with their light indications. They test the GFCI by creating a ground fault via safety ground. If ground is not there, then they can't work, and so them not working is expected and desired. If they worked on a groundless socket, that would be bad, it would suggest bootlegged grounds. Commented Jun 1, 2021 at 7:17

3 Answers 3


The open ground is because there are only 2 conductors not 3 this is normal for your condition and a GFCI is required to protect the circuit to allow 3 prong receptacles down stream, note use the GFCI protected no equipment ground if protecting downstream 2 wire receptacles that you are changing to 3 prong/ wire.

Note the receptacles that are 2 wire and labeled will not trip the GFCI if the test button is pressed because there is no ground but are protected because of the electronics in the GFCI receptacle or breaker.


TL;DR That is 100% correct functionality.

A magic 8-ball tester is really just 3 lights. One between hot & neutral (expected to be on for a good circuit), one between hot & ground (expected to be on for a good circuit) and one between neutral & ground (expected to be off for a good circuit).

Since you don't have a ground wire connected, there is no current flowing between hot & ground, so that light is off which is interpreted (correctly) as "open ground".

When you press the GFCI test button, what you are actually doing is sending some current from hot to ground. That is normally not a good idea, which is why we have GFCI in the first place! If everything were connected properly, then that would result in the GFCI tripping. In that situation, if the GFCI is functioning properly, power to hot would be stopped. That would result in no lights on very quickly. So the lights while pressing the GFCI test button are undefined and meaningless - and normally not visible.

Since you don't have ground connected, while you are pressing the button you have:

  • Hot to neutral - since the switch has no reason to interrupt this.
  • Hot to ground - since internal to the tester hot is connected to ground (deliberately) so the light showing hot to ground turns on.

This is equivalent to a hot/neutral reverse. In a hot/neutral reverse, the hot to neutral light is still on (120V between them, just "opposite" but with AC that makes no difference) and neutral to ground is on because neutral and ground are 120V apart. In this case, hot is being sent to ground, so that makes ground 120V apart from neutral. The end result is that, technically speaking, when you press that button you are simulating a hot/neutral reverse, even though what you are really trying to do is create a low-level ground fault.


I experienced this when there was no main bonding jumper at the first point of disconnect. As soon as one was added, the GFCI worked and tripped as expected with a plug-in tester.

Not the answer you're looking for? Browse other questions tagged or ask your own question.