Psychological & Neuropsychological Testing Billing (CPT 96130-96139): Recover the Denied Battery

By Rcmaxis Revenue Cycle Team  ·  September 25, 2026  ·  10 min read

A neuropsychologist finishes a full battery on Friday. Interview, six instruments, scoring, a 14-page report, a feedback session. Roughly nine hours of work. The claim goes out with 96132, one unit of 96133, 96136 and a few units of 96137, and comes back three weeks later with the evaluation lines denied and the administration lines paid at the wrong level.

Nobody made a coding mistake in the classic sense. The clinician did the work. The chart just can't prove it in the format the payer wants. If your practice runs 40 batteries a month and a hypothetical $500 average allowed amount per case, then a 10 percent problem rate is $2,000 a month, or $24,000 a year, before you count what it costs to chase the appeals.

Psychological testing billing (CPT 96130 through 96139) is one of the most documentation-sensitive lines in behavioral health. It rewards practices that keep a clean time log, and it punishes everyone else.

The Root Operational Friction: Why Testing Claims Fail

Testing isn't one service. It's three, stacked, each with its own clock and its own rendering person:

1. Evaluation services (96130/96131 for psychological, 96132/96133 for neuropsychological): the physician or qualified health care professional integrates data, interprets, makes clinical decisions, writes the report and gives feedback. Billed by the hour.
2. Test administration and scoring by the professional (96136 first 30 minutes, +96137 each additional 30).
3. Test administration and scoring by a technician (96138 first 30 minutes, +96139 each additional 30).

Then there's 96146 for a single automated instrument administered by computer with no professional time. Add 96127 for brief emotional or behavioral screens scored per instrument. That's six different billing behaviors in one chart.

The friction shows up in a few predictable places:

• The time isn't logged by activity. A single "testing: 6 hours" line is a denial waiting to happen. Payers want dates and start and stop times for each activity.
• Who did the work isn't clear. If the technician gave the tests and the psychologist scored them, the split is real and it needs to be on the record. If it isn't, the biller defaults to whatever's easiest.
• The interview gets counted twice. Clinical interview time belongs inside the evaluation code. If the same 50 minutes shows up under 90791 and again in 96130, NCCI edits and reviewers will catch it.
• The time rules get ignored. Time units are met at the midpoint. A "first hour" code needs more than 30 minutes of the work in that block, and each 30-minute administration unit needs more than 15. Billing a unit for 12 minutes is a compliance problem, not a rounding choice.
• Medical necessity is assumed. The referral question, the reason these specific tests, and the diagnosis that supports coverage have to appear in the record before the claim, not after the denial.

Specific Payer Edits, Modifiers, and Coding Traps

Code family at a glance

The traps that cost real money

1. One person, two administration codes. 96136/96137 and 96138/96139 describe who did the administering and scoring. A given block of time belongs to one or the other. When a psychologist and a technician each work on the battery, both can be billed, but only for their own time, and the log has to show it.

2. Technician supervision language. Medicare and most payers expect the technician to work under the supervision of the billing professional, and the level required can differ by payer and by state. Document who supervised, where they were, and confirm your MAC's current requirement instead of relying on the last policy you read.

3. Interview plus evaluation double-dipping. NCCI expects the interview and the evaluation to be distinct, without the same time in both. If you bill 90791 or 90792 on the same day as 96130 or 96132 for the same patient, look at your NCCI edits and your notes. One of the two probably has time it shouldn't.

4. Psychotherapy on the same day. Billing 90832 to 90838 in the same session block as testing feedback is a favorite reviewer target. If therapy and testing happened on the same day, the notes need clearly separate times and separate clinical purposes.

5. Repeat testing with no change in condition. MAC articles typically expect a documented reason for testing that repeats within a defined period, such as a change in clinical status or a new referral question. "Follow-up battery" isn't one.

6. Exceeding MUE caps. Long neuropsychological batteries can push add-on units past the Medically Unlikely Edit for a line. Pull the current MUE for each add-on code, and have the psychologist document why the higher-unit service was reasonable and necessary before you append a modifier and appeal.

7. Medical necessity gaps. The neuropsychological codes (96132, 96133, 96136 to 96139 when used for neurocognitive testing) are aimed at diagnosing and characterizing the neurocognitive effects of medical conditions that affect the brain. Testing for educational placement or work-related evaluations isn't a covered use under most Medicare articles. The record has to tie the tests to a diagnosis from the payer's covered list.

Prior auth and payer rules: what to actually do

Original Medicare doesn't require prior authorization for these codes, but Medicare Administrative Contractors publish billing and coding articles with covered diagnoses and frequency expectations (see the CMS article on psychological and neuropsychological testing, A57481, and your own MAC's LCD). Many commercial and Medicare Advantage plans do require authorization and cap units per request. Here's a practical playbook:

1. Ask for units by code up front. Request the full battery in one submission. A typical request lists 96130 x1, 96131 x2, 96136 x1, 96137 x4 (or the 96138/96139 equivalents), with the hours by activity. Extensions are slower and harder than a right-sized first request.
2. Send the request package that gets approved. Referral question, relevant diagnoses, prior treatment and results, planned instruments and why each one is needed, and the unit estimate by code.
3. If the request is denied or cut, ask for the clinical criteria used, then request a peer-to-peer with the psychologist, not the billing office. Appeal in writing with the referral question and the specific instruments tied to the diagnosis.
4. Track authorizations by date and units used, not by "authorized: yes." The authorization has an expiration date and a unit ceiling. Log both in a worklist, and check units remaining before the last testing session, not after.

The Tactical Workflow Fix: Step by Step

You can close most of this gap in a month with templates and one checklist.

Step 1. Build a testing log template. One row per activity. Columns: date, activity (interview, test administration, scoring, interpretation, report writing, feedback), who did it, start time, stop time, minutes. Attach it to every battery.

Step 2. Set the chargemaster to match the work. Load 96130 to 96139, 96146 and 96127 with descriptors that match CPT. Add a rule that blocks 96136/96137 and 96138/96139 from being billed for the same time block.

Step 3. Add a pre-bill checklist. Six questions before release: Is the referral question in the record? Does the diagnosis support coverage? Is the log complete and totals match units? Did a technician work under documented supervision? Is the interview counted once? Are authorization units on hand?

Step 4. Round by rule, not by feel. Use CPT's time rule: count a unit when the midpoint is passed. Build the calculation into the template so billing doesn't need to do arithmetic on the fly.

Step 5. Review denials weekly. Sort by payer and reason. If the same payer denies the same line twice, it's a payer edit you haven't documented yet. Send it to your denial management workflow.

Step 6. Audit five charts a month. Pick five batteries, rebuild the claim from the log, and compare to what was billed. Any variance goes to the psychologist and the biller together.

Revenue Impact: Days in AR and Collection Benchmarks

• Recovered revenue. A practice running 40 batteries a month with a $500 average allowed amount (a planning illustration, not a benchmark) loses about $24,000 a year at a 10 percent problem rate. Cutting that in half recovers about $12,000 and removes most of the appeal labor.
• Days in AR. Testing claims are high-dollar and slow when they go wrong. A healthy target is under 40 days in AR for this line. Above 55 days usually means claims are stuck in documentation requests or authorization disputes.
• Net collection rate. 95 percent or better of allowed amounts is a fair benchmark. Testing practices that sit near 85 percent almost always have unposted denials or write-offs that should've been appealed.
• First-pass clean claim rate. Above 92 to 95 percent is realistic once the log template and checklist are in place.
• Appeal effort. A denied battery can eat two to three hours of staff time. Prevent it and that time goes back to follow-up on claims that are actually aging.

For how these fit with the rest of a behavioral health revenue cycle, see our behavioral health billing guide, our neurology billing guide and our psychiatry billing page.

OIG, RAC and MAC Audit Exposure for Testing Services

Testing is high-dollar per claim and time-based, which is the profile reviewers like.

• CMS error-rate reporting. CMS's CERT program (the annual improper payment measurement) publishes improper payment rates by service category, and documentation-related errors dominate. Time-based services with weak logs land in that bucket.
• OIG. OIG audits of Medicare behavioral health and psychotherapy payments are a recurring theme, and the OIG Work Plan is where new topics appear. Check it quarterly for behavioral health items.
• MAC review. MACs use billing and coding articles as the standard for medical necessity. If your diagnosis isn't on the covered list for the code you billed, the claim fails on that alone.
• Typical error patterns: units billed without a supporting time log, technician services billed as professional services, testing that doesn't tie to a covered diagnosis, and repeat testing with no documented change.
• Lookback. RACs generally look back up to three years. The 60-day overpayment rule (42 CFR 401.305) has a six-year lookback for reporting and returning identified overpayments.
• What to do now. Pull the last 12 months of 96130 to 96139 claims by provider. Compare units billed to minutes logged for ten claims. If the numbers don't match, that's your audit finding, and it's better to find it yourself.

Frequently Asked Questions About Psychological Testing Billing

What's the difference between 96130 and 96132?

96130 (with add-on 96131) is psychological testing evaluation. 96132 (with add-on 96133) is neuropsychological testing evaluation. The neuropsychological codes focus on the cognitive effects of conditions that affect the brain and typically carry stricter medical necessity expectations.

Can a technician's time be billed under 96136?

No. 96136 and 96137 are for administration and scoring by the physician or qualified health care professional. Technician time goes under 96138 and 96139, with supervision documented.

Can we bill 90791 and 96130 on the same day?

It can be done when the services are distinct and the time isn't counted twice, but it draws review. The interview belongs inside the evaluation code when it's part of the testing evaluation. Check your NCCI edits and document separate purposes and times.

How much time is needed to bill an additional unit?

CPT's midpoint rule applies: a unit is met when more than half of its time is spent. For a 60-minute unit that's at least 31 minutes, and for a 30-minute unit at least 16 minutes. Document start and stop times for every activity.