Why can’t you connect neutral to ground to convert an outlet from 2-prong to 3-prong?

Not everything I put in the riddles, puzzles, and brain teasers section involves a king and some sort of puzzle that has to be solved with fancy logic. Sometimes it’s a real life question, and this is one of those times.

The question of course is if you have a 2-prong outlet (live + neutral) and want to convert it to a 3-prong (live + neutral + ground), why can’t you simply use a little jumper-wire on the new outlet and hook up ground to neutral? After all, they both go to ground eventually. This would mean that your surge protector that needs ground to operate correctly would work (surge protectors simply dump surges to ground before they hit your equipment). The problem is, by connecting neutral to ground in the outlet, there’s a situation that could happen that could kill you. No, it doesn’t have to do with the surge protector or power surges. Want to guess what it is?


If the neutral wire in the walls were ever to become disconnected/broken/etc, any 3-prong devices that were connected to the outlet (for example, a computer) would now have a live case. If you touched the case, you could be electrocuted.

You see, the metal case on 3-prong devices is always connected to the ground prong, mainly because if there were a problem in the device and the case somehow became live, that power would have a clean path to ground and would hopefully blow a circuit breaker before any harm could be done.

Since you connected the neutral prong directly to the ground prong, the case itself now has the potential to carry the return current. If the neutral connection were ever broken, it would, and if you touched the case (and were yourself grounded), your body would be completing the circuit. The device might even power on for as long as you were touching it. Unfortunately, all that current would be running through your body. For those of you who have seen the “call before you dig” commercials, you’ve probably got an idea of what would happen next… Never a good thing.

Note for those who haven’t seen the “call before you dig” commercials…

You die.


Update: Adding an image to help visualize my explanation. I used a toaster so I could “draw” the electrical internals (a heating element in this case), but you can pretend it’s a computer or anything else with a metal case if you prefer.

(click for a larger version)

3 prong outlet jumpered neutral to ground example

Since my drawing skills aren’t exactly going to win any awards, to make it easier to follow start by identifying the:

  1. TOASTER, OUTLET (plug), and FUSE PANEL.
  2. WIRES

…note that the jumper wire you hopefully didn’t add is shown at the bottom-right of the outlet (plug).

If you remove the jumper wire, the flow of power stops here instead of being routed to the toasters case via the toasters ground wire. The toaster doesn’t work, but touching the case doesn’t zap you… unless of course you’re really unlucky and the toaster has an internal short to the case that has gone unnoticed up till now… at which point hopefully you have a working GFCI.

28 Comments | Leave a Comment

 Sort by Oldest | Sort by Newest
  1. Steve Gergetz on October 14, 2009 - click here to reply

    Thanks for a concise and very clear explanation. I've heard it before but couldn't remember the explanation. Your version of it was stated so simply I'm sure it will stick in my head this time.

  2. sky jumper on March 9, 2011 - click here to reply
    I guess the problem with this explanation (not just yours, but the electrical code as well) is that it requires BOTH a fault in the device that shorts the chassis to AC "hot" AND an open ground/neutral feed for the danger to exist. To me this seems very unlikely. Further, you'd have to assume that whatever caused the neutral to open circuit would not also cause the ground to open - since the wires are all run together I think it's likely that any disruptive event that would open the neutral (e.g. a contractor's recip saw) would also cut through the ground and hot wires. I'll admit that I'm not an electrical inspector or fire investigator that has seen many cases of improper grounds causing problems -- but the logic behind it still escapes me.
    • Matt Gadient on December 3, 2011 - click here to reply
      sky jumper: I'm about 9 months late on the response here (sorry).

      But no, it doesn't require a fault that shorts the chassis to hot - just the open neutral. A picture would probably have been the easiest way of showing this, but if the ground prong is wired to the neutral prong via a jumper and you get a broken neutral, essentially this is the route the power takes:
      -"hot" wire
      -"hot" prong
      -through device internals
      -neutral prong
      -"jumper wire" you put in
      -ground prong

      The moment you touch the device-case, your body will complete the circuit if you're grounded.
    • Jalen on December 18, 2011 - click here to reply
      Deep thought! Thanks for contribtuing.
  3. Tyler on March 20, 2012 - click here to reply
    I'm a mechanical person without much electrical experience, so please correct me if my thinking is off. If you put a jumper between ground and neutral, and the neutral line somehow is disconnected, is that not the same as the neutral connection failing in a two prong outlet? Either way, you could be shocked? Or is there some extra protection within electronics with the two prong? I appreciate the help.
    • Tyler:

      The easiest way to explain this is probably to have you take a gander at the Wikipedia entry on Appliance Classes http://en.wikipedia.org/wiki/Appliance_classes .

      Basically, 2-prong devices are supposed to be designed in such a way that a single failure inside the case won't bring dangerous voltage to the conductive casing. If the neutral becomes disconnected, the device just plain shouldn't work. You shouldn't receive a shock by touching the case because the case should be well-insulated from any high-voltage internals. Or... they might have just used a plastic casing which won't conduct electricity anyway. Either way, the case should always be "dead" in a 2-prong device unless you have multiple failures going on within.
  4. Internetgent on January 9, 2018 - click here to reply
    This whole thread was exactly what I needed to know. Thank you very much for getting on the internet
  5. Anonymous on February 21, 2018 - click here to reply
    The ground and neutral have a bridge in the breaker box, so what's the concern
    • Matt Gadient on February 21, 2018 - click here to reply
      I just added an image to help show the issue.

      However, the main differences between joining-in-panel vs joining-at-outlet are that:
      • A lot of extra points of failure doing it after all the household (wall) wiring at an outlet.
      • The bonding in panels doesn't just bond your household outlet end... it also bonds the incoming neutral (from carrier) and ground rod. The panel has effectively has 2 really good paths to ground. The "equivalent" at an outlet would be bonding neutral and ground in a 3-wire incoming household feed (L+N+G). Though you're not allowed to do it, and it wouldn't really make sense anyway.
      • Panel failures are much less likely than any household wiring after the panel. Panel failures are also much more likely to be really catastrophic. The chance that a panel failure results in a shock when touching a toaster is quite a bit lower than the chance a panel failure results in a big spark show as the opening event to burning a house down.
      Really, jumping live to neutral to convert a 2-prong outlet to 3-prong is just not at all a good idea.

      If wanting a working ground for safety reasons, you're better off using a GFCI (if you don't have a real ground wire, just don't connect anything to the ground screw - only connect live + neutral). The GFCI will generally trip (and save you from shocks) if the device shorts to the case internally, falls in a sink full of water, etc. Using a GFCI outlet in this fashion (to upgrade a 2-prong outlet) is allowed by the electrical code in a number of areas.

      If wanting a working ground for what I'll call "electronics" reasons (electrical noise, power bar warning lights, surge protection, etc), really your only good option is to run some new 3-wire (L+N+G, often 14/2 romex in US/Canada for a 15A circuit). If that absolutely isn't an option, you could feasibly see if the metal outlet box is grounded. Grounding to the box when you don't know *how* the box is grounded isn't exactly a great idea since whatever is grounding it might-not-be-capable-of-carrying-15A-to-ground and hotspots/fires become a risk if there's a short in the device and the conductor isn't up to the task. Examples:
      • Old BX cable (metal clad/shielded without a ground strip) used the spiral metal shielding as the ground. But the metal shielding has higher resistance (it's not a good conductor like copper or aluminum), and thus can glow red if it's actually asked to pass a lot of current, especially in rusty sections, which could start a fire. It's no longer allowed in new installations by most electrical code. Good news is that it probably wouldn't have trouble handling the current of typical electrical noise. Bad news is the resistance is so high it might not be great at reducing electrical noise much anyway.
      • Properly gauged, properly grounded copper wire has sometimes been fished through walls/ceilings to each box to retrofit existing 2 wire. This is probably the best case scenario since it's a good conductor.
      • A small run of metal clad cable that isn't grounded might be rubbing up against something that actually *is* grounded (water pipe etc). Bad scenario because this is not a solid, reliable ground.
      • A small run of un-grounded metal clad cable might be joining 2 or more outlets. One of the things plugged into that outlet might have it's case grounded (metal case on concrete basement floor for example) and is proving the "ground". This is quite a bad scenerio.
      Even the "intended" grounds in the list may not be code-approved in your area. And it's not always easy to figure out what's actually grounding your box. But take it for what it's worth.
  6. Bill on July 12, 2018 - click here to reply
    it's all an safety concern. it's to eliminate the possibility of it happening. if their is just a chance of it happening aka [accident] its better to fault on the safe side. then take the chance. not saying it would happen. but the potential for it happening is their. that' s why. safety first. products are in our house from all over the country. god knows how they are wired. or if inspected. to late after it happens to say op ps. and the danger of a person doing his own wiring. and then selling the house. and someone else buys and don't know what he did. makes it dangerous. go the extra mile to be safe. its time well spent. Bill
  7. Ken on January 14, 2019 - click here to reply
    Hi, I'm a noob on this field and I'm just curious. The smart wall switch I bought off an online store required a neutral but our electrical circuit setup here in our home only has two. We only have the hot wire and the ground wire. Our electrician managed to have the switch working by having one end of the wire connected to the neutral field of the smart wall switch and the other end to a metal planted in the ground soil. Is it by any means dangerous? Thanks in advance.
    • Matt Gadient on January 14, 2019 - click here to reply
      Hard to know for certain without looking at and verifying the setup. Would also need to know something about the switch itself (some stuff bought online can be rather dubious in quality and design). Would also need to know something about the power distribution in the area... not having a neutral makes me wonder if the distribution in your area might be single-wire earth-return.

      In any case, generally an electrician will do things according to local code. If uncertain, you may want to consider running it by another electrician in your area.
      • Ken on January 15, 2019 - click here to reply
        Thanks for your response Matt. This is the exact unit I bought online -
        • Matt Gadient on January 15, 2019 - click here to reply
          I took a look at the link to the product you mentioned. Their diagram showed just Hot/Neutral (no separate ground). So if in your country Hot/Ground is standard house wiring (as opposed to 3-wire Hot/Neutral/Ground), my initial thought would be that Hot/Ground might be treated as what we'd see here as Hot/Neutral. However if that were the case I'm not sure why your electrician would have run a separate ground plate for this installation.

          That said, I could easily be missing something, and it's very possible that things are done very differently in your country. Again if unsure I'd really consider running it by an electrician there.
  8. Anonymous on January 23, 2020 - click here to reply
    well i have touched 120 volts lots of times. its just mildly annoying. i suppose if conditions were just perfect it could kill you but i think its extremely unlikely.
  9. Debo on April 9, 2020 - click here to reply
    I live in an old home and the outlets that I thought were properly grounded, I later found out they just connected the neutral and grounding wires just behind the outlet. One electrician seems to think its not a big deal, another seems to think my house will burn down.
    Can I just change all the outlets to gfci and be done with it?
    • Hey Debo,
      • GFCI + keeping the ground/neutral jumper: Adds the standard GFI shock protections (5-20mA), does not protect against the specific failure mode above (it occurs post-GFCI).
      • GFCI + removing the ground/neutral jumper: Similar to a 2-wire to GFCI upgrade and is the "approved" method of converting a 2-prong outlet to 3-prong outlet in some places. Some GFCIs even come with "not grounded" stickers that can be placed on the wall plate. Provides GFCI shock protecion (5-20mA). Removing the jumper mitigates the specific failure mode above.
      I should be clear that removing the neutral/ground jumper does present a trade-off. An unconnected ground has its own set of possible issues: some devices really like to be grounded. GFCIs also don't protect against painful high voltage shocks if they occur at low current - there's a current threshold that has to be met (roughly 5-20mA of current as I just mentioned). If you have 3 prong devices that actually *do* cause the case to become live at a low current but normally dump it to ground... remove the jumper wire and you might find you're getting some new shocks or fuzzy-feels when you touch the device whether or not a GFCI is involved (though the GFCI should trip before the current gets to typically lethal levels).

      As to the house burning down... not sure if there were other factors at play..? The jumped ground/neutral on its own shouldn't have any bearing on fire risk unless there is something wonky about it (undersized, not well fastened, resulted in too many wires for the box, other stuff that applies to all electrical installations, etc). The risk is shock if the neutral in the wall becomes disconnected and a person touches the case while they are grounded. It has nothing to do with fire.
      • Rudy on September 7, 2020 - click here to reply
        Since we have a code violation for an unconnected ground to our 3-prong outlets in the house, rather than replace with all 2-prong outlets, I would prefer to install GFCI outlets. Do I have to replace *all* of the outlets with GFCI, or do I only have to replace the first GFCI outlet in that particular leg of the power (as we do with properly grounded systems)? I am thinking I would need to replace them all, due to no ground connection. But, wanted to check that beforehand.

        This is an approved method per the NEC, isn't it? (We would be using the "not grounded" stickers, of course.) We just don't want the inspector coming back on us to replace all the GFCIs with 2-prong outlets. (I realize the ground isn't connected, but I feel it's safer than the cheater plugs dangling from the walls!)
        • Matt Gadient on September 7, 2020 - click here to reply
          Practically speaking, a GFCI doesn't rely on the on the ground for operation, so using a GFCI on the 1st outlet in the leg would protect all the others downstream (providing LINE heads towards the panel and LOAD heads to the downstream receptacles).

          However, local codes can vary, and from what I gather in some locations inspectors have been known to refuse certain things that they don't like despite meeting the code. So you're really best to talk to a local inspector or electrician.
  10. Troublemaker999 on July 13, 2021 - click here to reply
    The only flaw in your argument is that the neutral and ground are bonded together in the service panel.
    • Indeed. Bonded in at the panel, but not permitted to be bonded at the outlet. Examining possible failure modes shows why this is the case!
  11. Ronen on July 23, 2021 - click here to reply
    So the whole purpose of the third ground wire is to prevent a shock when the hot wire accidently touches the case and the neutral wire is accidently broken and disconnected from the device, and the person is touching the case bare footed on a conductive floor or is touching a conductive wall or is touching another appliance at the same time that its neutral wire accidently touching the case.
    What if the ground cable also breaks? Maybe we should add more ground cables to lower this close to zero risk?
    This problem could easily be solved anyway by not grounding the neutral, and keep it hot.
  12. Anonymous on September 2, 2021 - click here to reply
    To the old house... you need to run a wire from ground to all the outlet circuits and attach it to the neutral at the load center to ground the neutral... because even though the neutral and ground are attached at the load center, you need that third wire that you are missing at the outlet connections to be independent of the neutral! This way you don't overload the neutral... (the same wire diameter that you use in the phase, you must use in the neutral, and also in the ground).
  13. Oscar on September 7, 2021 - click here to reply
    Best answer, very respectfull and concise. In the whole internet. And great drawing; that perfected my comprehension.
  14. Tamer on November 12, 2021 - click here to reply
    Thanks very much for the topic discussion & sharing...
    Here the issue:
    i don't have ground wire at all, and my PC case always have a current in it, if ever i was bare footed & touch the case i got shocked, we have 220V not 110V, also i recently bought a laser machine which needed to be grounded.

    after some search online i came up with a solution, not permanent, but it did the job.

    i made a jumper between the neutral & the ground at the wall outlet and it worked like a charm!
    before i took that step i asked many electricians in my circle if i can do this, but i got surprised they don't know anything about that!

    of course this way provides a solution for devices already have a leakage, and work as a temp solution for devices which needs a ground connection in situations like mine, no grounding infra structure at all.

    Thanks again for the brainstorming :)
  15. JimS on December 20, 2021 - click here to reply
    Picture tells a thousand words, didn't understand until I saw it, very nice. Bad idea to jumper ground to neutral at the GFCI receptacle, but how about jumping ground to neutral on downstream 2-wire non-GFCI receptacles that are protected by an upstream GFCI device?
  16. Dave on January 18, 2022 - click here to reply
    Matt Gadient: Thank you very much. If the outlet was done connecting neutral to ground, would a 3-prong testing device show it was grounded and not recognize it was done creating a hazard?
    • Matt Gadient on January 18, 2022 - click here to reply
      I haven't verified this, but my suspicion is that a typical tester would indicate that the circuit was grounded. The reason is that neutral and ground should be at the same potential and both are bonded at the panel - essentially a tester would have to distinguish between bonding at the panel vs bonding at the outlet. There are costly ways to take a shot at it (signal distance and/or using resistance as a hint) but I'm not sure if there's an easy and cost-effective way to do that. Maybe there is and I'm just not aware of it though.

Leave a Comment

You can use an alias and fake email. However, if you choose to use a real email, "gravatars" are supported. You can check the privacy policy for more details.