National Psychiatry Supply by State — NPPES 2026 Snapshot

Mental health workforce shortages are one of the most-cited public-health gaps in U.S. medicine. The U.S. Health Resources and Services Administration (HRSA) maintains a federal Mental Health Professional Shortage Area (HPSA) designation; as of HRSA's 2024 dashboard, approximately 158 million Americans live in a designated mental health workforce shortage area. KFF's 2025 Mental Health Care Workforce tracker and the SAMHSA 2024 National Survey of Substance Use and Mental Health both flag a structural under-supply of psychiatrists relative to demand.

This study reports what the public NPPES dataset shows: every active NPI-1 psychiatrist (NUCC taxonomy 2084P0800X) bucketed by Practice Address state and divided by U.S. Census Bureau 2024-vintage state population. It is a density measurement at state granularity — not a spatial-access model, not an outcome study, and not a count of all mental-health providers. PAs, psychiatric nurse practitioners (PMHNPs), psychologists, and licensed clinical social workers each carry their own NPPES taxonomies and are out of scope here. They matter — they expand mental-health capacity materially — but they are not psychiatrists, and conflating the two obscures the supply picture this study is designed to surface.

Source. The U.S. Centers for Medicare & Medicaid Services (CMS) National Plan and Provider Enumeration System (NPPES) — the public NPI Registry. Every U.S. healthcare provider who bills any payer (commercial, Medicare, Medicaid) holds an NPI; NPPES is the registry of record for that identifier. The public API at https://npiregistry.cms.hhs.gov/api/?version=2.1 requires no authentication and is updated continuously by CMS as providers self-attest changes.

Counted (kept). Each provider in our snapshot meets all three criteria:

1. Type 1 NPI (individual practitioner). Type 2 NPIs (organizations / group practices) are a separate analysis surface and are not counted here.

2. Active — basic.status = "A" and basic.deactivation_date is null.

3. At least one Healthcare Provider Taxonomy code in the psychiatry family:

   • 2084P0800X — Psychiatry (individual provider)

This study counts only physician psychiatrists under NUCC taxonomy 2084P0800X. Psychiatric subspecialties (Child & Adolescent Psychiatry 2084P0804X, Geriatric Psychiatry 2084P0805X, Addiction Psychiatry 2084P0802X, Forensic Psychiatry 2084F0202X, Psychosomatic Medicine 2084P0901X) carry their own codes and are out of scope. PMHNPs, psychologists (103T*), licensed clinical social workers (104100000X), licensed counselors (101Y*), and marriage & family therapists (106H*) — all material to mental-health capacity — are out of scope and would be a separate study.

State assignment. Each NPI's Practice Address (LOCATION) state is used, not the Mailing Address. NPPES distinguishes the two; for cross-state telemedicine practices the Practice Address state may differ from where the provider primarily lives.

Not counted. Providers with deactivated NPIs, providers whose only psychiatry-adjacent taxonomy is outside the codes above, residents and fellows whose primary taxonomy is the "Student" code (390200000X), and providers without a U.S. Practice Address.

The national rate of 18.26 active psychiatrists per 100,000 residents is a single number; the state spread is what consumers, providers, and policy readers actually feel.

Top 5 by density:

• District of Columbia — 71.34 / 100k (n=501)
• Massachusetts — 41.34 / 100k (n=2,950)
• Rhode Island — 37.94 / 100k (n=422)
• Connecticut — 35.59 / 100k (n=1,308)
• New York — 35.02 / 100k (n=6,957)

Bottom 5 by density:

• Wyoming — 7.83 / 100k (n=46)
• Idaho — 9.09 / 100k (n=182)
• Mississippi — 9.85 / 100k (n=290)
• Nevada — 10.56 / 100k (n=345)
• Alabama — 10.59 / 100k (n=546)

The lowest-density state holds approximately 11% of the highest-density jurisdiction's per-capita supply. The chart below shows the full ranking; the dashed marker is the 10-per-100k threshold this study uses to flag underserved jurisdictions.

3 jurisdictions rank below the 10 active psychiatrists per 100,000 residents threshold this study uses to flag underserved jurisdictions. The threshold is a transparent baseline cutoff stated explicitly here — it is not a clinical or regulatory definition. Aligned with HRSA Mental Health Professional Shortage Area (HPSA) criteria, which use a population-to-psychiatrist ratio of approximately 30,000:1 to designate shortage areas — equivalent to roughly 3 psychiatrists per 100,000 in the most restrictive HPSA cutoff, with workforce-adequacy thresholds in the 8–12-per-100k range cited across HRSA and KFF mental-health workforce trackers. We use 10 / 100k as a transparent baseline cutoff that sits squarely in that adequacy range. Stated explicitly here; not a clinical or regulatory definition.

The states below the threshold collectively hold 5,532,282 residents — about 1.6% of the U.S. population.

Important framing. "Underserved" here means density-below-threshold in the public NPPES dataset. It does not measure new-patient availability, insurance acceptance, sub-specialty match, travel time to the nearest practice, or whether the practice is accepting Medicaid. Network-distance and gravity-model methodologies produce a different, complementary picture — one we explicitly cite below.

Density per 100,000 residents is the most transparent supply baseline. It uses two public datasets, requires no modeling assumptions beyond arithmetic, and can be audited row-by-row in the downloadable dataset. Every reader can audit the numbers themselves.

It is not spatial access. A patient in a rural state and a patient in a metropolitan state both contribute the same "1" to their state's denominator, but their realized access differs by orders of magnitude. Headline figures derived from network-distance models (e.g. "X% of Americans live within Y minutes of a [specialist]") coexist with the present study's state-level density findings — both can be true simultaneously, because density and proximity capture different facets of "access."

The healthcare workforce literature has more advanced approaches:

• Two-Step Floating Catchment Area (E2SFCA) and gravity models treat access as a function of provider supply, population demand, and distance decay. A 2023 review of gravity models for potential spatial healthcare access (open-access spatial-epidemiology literature) is the standard reference.
• Network-based access measures compute drive-time isochrones from each Census tract centroid and count reachable providers within a window.

This study does not implement E2SFCA, gravity, or network distance. It reports state-level density and acknowledges the gap. Readers who need spatial-access estimates should pair our state ranking with the cited literature or specialty-specific access modules in the AAMC State Physician Workforce Data Reports.

Public-health relevance. Density of psychiatrists has been studied as a covariate of population mental-health outcomes. A 2024 Health Affairs paper on geographic variation in suicide rates and psychiatrist supply (and the 2023 JAMA Psychiatry literature on access barriers) finds that lower density correlates with poorer help-seeking outcomes, particularly in rural counties. The shape of the relationship matters: in states already at top-decile supply, adding more psychiatrists yields diminishing returns; in bottom-quartile states, the marginal value of one additional psychiatrist is substantially higher, because a county with zero psychiatrists is qualitatively different from one with even modest supply.

The bottom-quartile states in this study are the supply-side cohort where the mental-health workforce literature suggests capacity expansion has the highest marginal value. We are not claiming a causal link between any specific state's NPPES count and any specific outcome metric; the cited papers make those arguments with the appropriate covariate adjustment, and we link to them for readers who want the full chain.

Capacity expansion vehicles flagged in the literature: telepsychiatry (post-pandemic, geography-blind delivery), Collaborative Care Model integration in primary care, and the rapid growth of Psychiatric Mental Health Nurse Practitioners (PMHNPs) — none of which alters the NPI-1 psychiatrist count this study tracks. Future versions of the study may add a parallel track for PMHNP density (NUCC taxonomy 364SP0808X / 363LP0808X) once that is methodologically defensible to surface.

Citation transparency. This study makes no clinical claims about any individual provider. The density-to-outcomes link is one published in the public-health literature; we cite it as context for why density matters, not as a finding our dataset proves.

Most state-level psychiatry supply numbers in circulation come from one of three places: (1) the AAMC State Physician Workforce reports, which use specialty-society membership and AMA Masterfile cross-references; (2) commercial workforce-data vendors that license those underlying datasets; or (3) press-release figures from advocacy organizations.

This study is different in three concrete ways:

1. Single public source family. Everything reported here ties back to the public NPPES API and the public Census Bureau population estimate. Both are free. Both are updated on cadences CMS and Census publish. Anyone with a web browser can re-run the underlying queries.
2. Record-level traceability. Every provider in the count has an NPI. The downloadable dataset preserves NPI numbers, the matched state, and the per-state aggregate. A reader auditing a specific state's count can pull the corresponding NPI list, look each one up in NPPES, and audit the active-psychiatry-taxonomy criterion themselves.
3. No quality attestation. Fonteum does not run a checking process for individual psychiatrists. The label "active in NPPES" describes a CMS registry status, not a quality, board-certification, or insurance-acceptance attestation. Patients who need clinical confirmation should consult the American Board of Medical Specialties (ABMS) registry for board-certification status and their insurer's directory for in-network availability.

The downloadable CSV at the top of this study includes per-state count, density, rank, quartile, and the underserved-threshold flag. The downloadable JSON adds the snapshot metadata and source URLs.

Suggested citation:

Ownlisted Research. (2026). National Psychiatry Supply by State — NPPES 2026 Snapshot. Ownlisted. Retrieved from https://fonteum.com/research/psychiatry-provider-supply-by-state-2026

Reuse and attribution. Charts, tables, and the downloadable CSV / JSON may be cited or reproduced with attribution to Ownlisted Research and a link to this study. Carry the snapshot date (2026-05-06) so readers know the dataset version. Future NPPES snapshots will produce different state-level counts as providers update their NPPES status.

The methodology, the explicit underserved-threshold definition, and the density-vs-spatial-access distinction must travel with the figures. Per-state counts published without those caveats risk being misread as a clinical-access measurement, which the dataset does not support.

Press / media inquiries. Reach the Ownlisted Research team via the brand-hub contact page. We are happy to clarify methodology for health-policy and access-equity reporters; we will not provide patient-side clinical commentary.

Cited literature (suggested for follow-up reading):

• U.S. Health Resources and Services Administration (HRSA). Designated Mental Health Professional Shortage Areas (HPSAs), 2024 dashboard. — Defines the HPSA criteria and reports the ~158M-Americans-in-shortage-area figure.
• SAMHSA. 2024 National Survey on Drug Use and Health. — Mental-health utilization and unmet-need national-level statistics.
• Kaiser Family Foundation. Mental Health Care Health Workforce Shortage tracker, 2025. — State-level workforce-shortage benchmarking.
• Health Affairs / JAMA Psychiatry recent peer-reviewed literature. Density-to-outcome linkages; the most-cited 2024 paper documents geographic variation in psychiatrist supply against population mental-health outcomes.
• Gravity models for potential spatial healthcare access (2023). — Methodological reference for E2SFCA and gravity-decay approaches not implemented here.

• Density is not spatial access. Per-100,000 resident density measures supply-to-population ratio at the state level. It does not measure travel time, new-patient availability, insurance acceptance, or sub-specialty match. Network-based and gravity-decay methods (E2SFCA) provide more accurate access measurements. We have not implemented those methods.
• NPPES taxonomies are self-reported. Providers select their own taxonomy codes when registering or updating their NPI. There is no specialty-board cross-check built into NPPES; a board-certified physician and a non-board-certified provider who self-attests the specialty code both appear. Patients confirming clinical credentials should consult ABMS / AOA registries directly.
• Practice Address state, not where the provider lives or practices most. A multi-site practitioner is counted once at their Practice Address state. Cross-state telemedicine practices may be undercounted in the patient-facing state.
• Active-flag accuracy. NPPES updates active/deactivated status as providers self-attest. Some providers carry stale NPIs from earlier career stages. The deactivation flag captures most cases; minor over-counting at the margin is possible.
• Census ACS 2024 vintage uncertainty. State population denominators come from the Census Bureau 2024 V2024 estimate. ACS estimates carry margin-of-error bands; per-100k figures are point estimates from a point-estimate denominator and should not be treated as exact to two decimal places.
• No procedure-mix or sub-specialty breakout. This study does not distinguish sub-specialties beyond what the taxonomy codes encode. A provider's primary clinical practice may differ from their NPPES taxonomy code.
• Type 2 (organization) NPIs excluded. Hospital-employed psychiatrists who only operate under a Type 2 organization NPI are under-counted. Most psychiatrists carry both, so the under-count effect is small but non-zero.
• No outcomes / quality-of-care claims. This study reports NPPES-listed counts and Census-derived densities. It makes no claims about practice quality, board certification status, sub-specialty expertise, or any clinical outcome metric. This is not medical advice. Patients should look up board certification through ABMS or AOA, confirm insurance acceptance with the practice, and consult their primary care physician for referrals.
• Snapshot in time. Counts reflect the 2026-05-06 NPPES snapshot. The cached dataset is preserved at data/nppes/psychiatry-2026-05-06.json for re-analysis.