Meter Says Grounded, Tester Says Open: Bootleg Ground Decision Tree

Why this matters

A receptacle where your multimeter reads voltage from hot to ground (suggesting a grounded outlet) but a three-light receptacle tester or a more careful test calls out an open or false ground is a classic bootleg-ground trap. A bootleg ground is a jumper tying the receptacle's ground terminal to its neutral terminal, which fools simple meters and basic testers into reporting "grounded" because hot-to-bootleg-ground reads like hot-to-real-ground. But there is no actual equipment-grounding conductor back to the panel; the "ground" is just the neutral, which carries current and rises in voltage under load. This is dangerous: it energizes the equipment-grounding system and any bonded metal under fault or open-neutral conditions, and it defeats GFCI/equipment-fault protection. Reading the difference between a real ground and a bootleg is a life-safety call.

Symptom presentation

A two-light/multimeter quick check suggests the outlet is grounded (hot-to-ground reads near 120 V). But a proper ground-integrity test (a tester with a loaded ground check, a GFCI test that behaves wrong, or a measurement that shows the ground rising with neutral load) indicates the ground is open or bootlegged. The outlet may be a three-prong device on old two-wire (ungrounded) cable that someone "grounded" by jumpering to neutral. Appliances may shock, surge protectors may not function, and the customer may report a recent DIY outlet swap.

Quick checks

  • Two test methods disagree. A simple hot-to-ground voltmeter reads "grounded" while a loaded-ground tester or careful test reads open/false. Disagreement is the bootleg tell.
  • Reverse-polarity bootleg. A bootleg combined with a swapped hot/neutral can make basic testers read normal while the ground is actually live; treat any DIY-wired outlet with suspicion.
  • Cable type. Old two-wire (no separate ground) cable feeding a three-prong outlet is a prime bootleg candidate; inspect the cable in the box.
  • Ground rises with load. Measure neutral-to-ground; on a bootleg it reads near zero (because they are the same wire) yet ground-to-a-true-earth reference rises under neutral load.
  • GFCI behavior. A bootleg can cause a GFCI to behave abnormally or fail to protect; an EGC-dependent tester button may misreport.

Isolation tree

  1. Confirm the cable has a real EGC. Open the box (de-energized). If the cable is two-wire with no equipment-grounding conductor and the outlet is three-prong with a ground that goes only to the neutral terminal (or a jumper between neutral and ground screws), it is a bootleg. No real ground exists.
  2. Bootleg present? A jumper from neutral to ground at the device, or ground bonded to neutral anywhere downstream of the service, is a bootleg. The ground is electrically the neutral, so hot-to-ground reads like hot-to-neutral (about 120 V), fooling basic testers.
  3. Reverse-polarity bootleg (worst case). If hot and neutral are also swapped, the bootlegged ground can sit at line potential relative to true earth while basic testers read "correct." Test ground against an independent earth reference; a hot ground confirms this dangerous case.
  4. Compare to true earth. Reference the receptacle ground against a known-good earth/ground. A real ground tracks earth (near zero N-G drop unloaded). A bootleg ground rises with neutral load and may sit at neutral potential or, in reverse-polarity, at line potential.
  5. GFCI protection check. A bootleg/open ground means equipment-fault current has no proper return path; verify whether GFCI protection is present and functioning, since that is the only acceptable path for these outlets under code.

Confirming diagnosis

Confirm with a test that loads the ground or references true earth, because that is exactly what a bootleg cannot pass. A receptacle tester or method that draws a small current through the ground path will show the bootleg as open/false even though hot-to-ground reads 120 V. The definitive confirmation is opening the box and finding either no equipment-grounding conductor or a neutral-to-ground jumper at the device; the wiring proves it. Referencing the device ground to an independent earth and watching it rise with neutral load (or sit at line potential in the reverse-polarity case) also confirms there is no real ground. Two tests disagreeing (simple voltmeter says grounded, loaded/earth-referenced test says open) is itself near-diagnostic for a bootleg.

Remediation

  • Remove the bootleg jumper. Never leave a neutral-to-ground bond at a downstream device. Remove the jumper.
  • Provide a real ground, or comply via the code-allowed alternatives. Where no EGC exists, the code-compliant options for replacing a non-grounding receptacle are: a grounding receptacle only if an EGC is run/available, OR a GFCI-protected receptacle (or GFCI ahead of it) labeled "No Equipment Ground," OR a non-grounding receptacle. Choose the compliant path; do not fake a ground.
  • Reverse-polarity bootleg: correct hot/neutral immediately; this can leave grounds and enclosures energized.
  • After remediation, verify with a loaded-ground/earth-referenced test and confirm GFCI protection where used.

A bootleg ground, especially combined with reversed polarity, can leave the equipment-grounding system and bonded metal at line potential; this is an electrocution hazard. Never create or leave a neutral-to-ground bond at a receptacle. Verify de-energization before opening the box, and test grounds against a true earth reference, not just hot-to-ground voltage.

References

  • NFPA 70 (National Electrical Code) Article 250 (grounding and bonding), including the prohibition on downstream neutral-ground bonds (250.142) and equipment-grounding requirements.
  • NEC 406.4(D), Replacements, for the allowed methods of replacing non-grounding receptacles (grounding type with EGC, GFCI-protected with "No Equipment Ground" marking, or non-grounding type).
  • NEC Article 210.8 and 406 for GFCI protection and receptacle requirements.
  • NFPA 70E for energized-work safety and verification of de-energization.