EMG and nerve conduction study billing: the complete guide.

EMG and NCS carry neurology's densest bundling problem, and it isn't subtle once you know where to look: the entire needle-EMG code set forks into two mutually exclusive families depending on a single fact — was a nerve conduction study performed the same session — and billing the wrong family is the leading cause of a bundled-service denial in this specialty. This guide covers the full code family, the unit-count rule that decides whether a study is "complete," the per-nerve documentation standard payers actually check, and the current Medicare local coverage article for every MAC that publishes one.

The three code families at a glance

Everything in this guide sits under three code families, and the decision tree between them is simpler than most practices treat it: how many nerves were studied, whether a needle EMG was performed the same session, and if so, how many muscles it covered.

The families are mutually exclusive by design, not by payer preference: the add-on codes (95885–95887) require a primary NCS procedure on the same claim to be billable at all, while the standalone codes (95860–95864) exist specifically for encounters where no NCS was performed that day. A claim that reports a standalone code and an NCS code on the same date, same patient, is reporting the needle-exam work under a rule set that assumes no NCS happened — which is exactly the pattern that trips a bundling edit, covered in more detail below.

Nerve conduction studies: 95907-95913

NCS codes are selected by counting the individual nerve conduction studies performed — each motor or sensory nerve tested on a side counts as one study, so a bilateral median motor study is two studies, not one. The code tiers up from there.

Counting error is the most common charge-capture mistake in this family, not code selection error — a coder who correctly identifies "this is the NCS family" still under- or over-tiers the code if the study count from the report isn't tallied precisely. F-wave and H-reflex studies count toward the total the same as standard motor and sensory conduction studies; a report that performs median and ulnar motor studies bilaterally plus bilateral F-waves on those same nerves is at minimum six studies, not two, even though only two nerves were tested.

Standalone EMG vs. combined testing: the decision that drives most denials

This is the fork every EMG claim has to resolve correctly, and it isn't a coding preference — it's a fact about what was actually performed that session.

The fix is upstream, not at the claims-scrubber stage: whoever assigns the code needs to know, before selecting it, whether the electrodiagnostic study included nerve conduction testing that day. In practices where the same physician performs and interprets both components in one encounter, this is rarely ambiguous. It gets error-prone in group practices where EMG and NCS are sometimes split across visits or providers — confirm same-day status from the study report itself, not from a template default.

95886 unit-count rules: what "complete" actually requires

95886 is neurology's most misapplied EMG code, and the requirement is specific: five or more muscles studied in the extremity, innervated by at least three different nerves or arising from four different spinal levels, with NCS performed the same session. Fewer muscles than that — even with NCS same day — is 95885, the limited add-on code, not 95886.

The unit count follows muscle count and extremity count, not a flat per-limb charge — a bilateral complete study that meets the 95886 threshold in both legs reports two units of 95886, one per extremity, each independently supportable from the note. Billing 95886 because the study "felt complete" without the muscle names and count in the report is a chart that won't survive an audit, because the code's own definition is a documentation checklist, not a clinical impression.

The per-nerve documentation standard

Payers and auditors check specificity against the study report, not against the code billed, and this family fails audits on documentation gaps more often than on code selection itself.

• 1Name every nerve tested, by side. "Bilateral upper extremity NCS" doesn't support a code; "right and left median, ulnar, and radial motor and sensory NCS" does, and it's what a reviewer checks the study count against.
• 2Report the actual values, not just "abnormal" or "normal." Latency, amplitude, and conduction velocity for each nerve tested, with F-wave and H-reflex values listed separately when performed — these are what get counted toward the NCS tier.
• 3List the muscles studied on needle EMG by name, not by count alone. Required to support 95886's five-muscle, three-nerve (or four-level) threshold, and useful corroboration for any needle EMG code on audit.
• 4State the clinical indication driving the study. Radiculopathy, entrapment neuropathy, and generalized polyneuropathy point to different nerve and muscle selections; a report that doesn't tie the studied nerves back to the clinical question reads as a template rather than a targeted workup.

NCCI edit pairs in this family

Two CMS-published values decide whether a bundled or capped code in this family is billable at all: the NCCI modifier indicator on a Column 1/Column 2 pair (0 = never bypassable, 1 = bypassable with documentation, 9 = edit deleted), and the MUE Adjudication Indicator on a per-code unit cap (1 = claim-line edit, 2 = absolute date-of-service edit with no appeal path, 3 = appealable with documentation). Both are pair-specific and code-specific, and both change quarterly.

Where this shows up most in EMG/NCS billing: a standalone EMG code (95860–95864) reported against a same-day NCS on the same extremity is the pattern most likely to trigger a Column 1/Column 2 edit in this family, because the payer's system reads it as the needle-exam work reported twice under two incompatible rule sets — once as if no NCS happened, once as if it did. The add-on codes (95885–95887) carry their own per-code MUE caps tied to extremity and muscle-group count, which is the other place unit-count errors surface as a denial rather than an underpayment.

⚠️ On specificity: this build could not open CMS's primary NCCI Procedure-to-Procedure edit file or the MUE table directly to confirm the current modifier-indicator and MAI values for the specific EMG/NCS pairs named above — the same access limitation our other specialty guides on this site have hit against CMS's site during automated research. Treat the indicator and MAI framework above as confirmed CMS policy structure, but look up the current value for any specific pair in the CMS NCCI PTP Edits Lookup Tool before building it into a scrubber rule, because these values change quarterly and the lookup tool is the only source that's always current.

MAC-specific coverage: the Noridian retirement, and the other six

"The article says" means nothing in this family without naming the MAC and confirming the version hasn't been superseded. We confirmed the following directly against the CMS Coverage Database.

Seven MACs currently publish a billing-and-coding article for this family, and all seven titles are identical — "Billing and Coding: Nerve Conduction Studies and Electromyography" — which is exactly why citing by article number and contractor matters more than citing by title. Find your practice's MAC first, then pull that contractor's specific article, not a cached search result or last year's version, before building a coverage requirement into an order set or a scrubber rule. ⚠️ We confirmed the article IDs, contractors, and effective/retirement dates above directly against the CMS Coverage Database; the full covered-diagnosis lists and documentation requirements inside each article were not independently re-verified here (CMS's own article pages blocked automated retrieval during this build), so pull the specific article text yourself before finalizing an order-set rule against it.

For the payer's own authorization requirements on this family — not universal, but real under some commercial and Medicare Advantage plans — see Neurology prior authorization: MRI, EEG, EMG, and Botox. For the companion diagnostic-testing family that shares the same component-split and time-tiered billing logic, see EEG billing and coding: routine, ambulatory, and video monitoring.

Frequently asked questions

Related resources

• Neurology billing and coding guide
• Neurology prior authorization: MRI, EEG, EMG, and Botox
• EEG billing and coding: routine, ambulatory, and video monitoring
• All 25 specialties