What skilled-nursing facilities actually earn: the state of SNF margins

The median skilled-nursing facility posts a 9% EBITDAR margin and an operating line that sits at zero. Both numbers describe the same building. The gap between them is the whole job.

Skilled nursing looks like a healthy business until you put rent back on the page. Across 12,247 Medicare cost-report filings, the median facility runs a 9.2% EBITDAR margin. Capital and rent then take roughly eight points of revenue, and the median operating margin lands at 0.3%. Nearly half of all facilities finish the year underwater on an operating basis. The sector earns its keep before rent and gives most of it back after.

• Median revenue — $10.7M · Middle half $7.1M to $15.3M
• EBITDAR margin — 9.2% · Before rent; middle half 1.6% to 16.4%
• Operating margin — 0.3% · 48.7% of facilities are negative
• Occupancy — 83.8% · Back above pre-COVID levels

The headline margin is not the business. The spread around it is. Here are the sector numbers, from the filings, and what each means when sizing a target.

What the SNF sector looks like financially

The file holds 12,458 Medicare-certified SNFs across all 50 states, on FY2024 and FY2025 cost reports. After screening out consolidated chain filings and implausible outliers, 12,247 facility-level filings drive the numbers here.

Median annual revenue is about $10.7 million, with the middle half between $7.1 million and $15.3 million and the top decile above $21 million. The median facility is certified for 101 beds (middle half 72 to 130) and books roughly $107,000 in revenue per bed, or $358 per patient-day. These are mid-market businesses, intensely local: a 100-bed building serving one catchment area, not a regional platform. Revenue per day doubles as a read on acuity and payer mix: a facility near the $432 upper quartile is running more Medicare and skilled-rehab days than one at $307.

Ownership skews commercial: about 74% for-profit, 15% nonprofit, 11% government. That for-profit majority is why so much of the sector trades, and why cost-report benchmarks get used in live diligence. The facility count is also drifting down, with closures clustered in the low-occupancy, thin-margin tail and most exits for-profit buildings leaving on their own rather than being pushed out. That cuts both ways for a buyer: tightening bed supply can firm up occupancy and pricing in a constrained market, while the same distressed buildings are the cheapest entry point for an operator that thinks it can fix census.

Margins: what a typical SNF earns, and how wide the spread runs

Skilled nursing is measured on EBITDAR, not EBITDA, because most operators lease their real estate and the rent line moves with deal structure rather than the operation. Adding rent back puts an owned building and a REIT-leased one on the same footing. (For the mechanics, see what EBITDAR is and why post-acute uses it.)

The median facility runs a 9.2% EBITDAR margin. The middle half sits between 1.6% and 16.4%, and the tails run wider: the bottom decile posts a negative 7% margin, the top decile better than 22%. More than a third of facilities, 35.5%, come in under a 5% EBITDAR margin, the level most lenders treat as covenant-risk territory, and about one in five are negative outright. With the interquartile range spanning nearly fifteen points, a national median tells you almost nothing about a specific building.

Now put rent back. The median operating margin is 0.3%, capital and rent absorb a median 8.2% of revenue, and 48.7% of facilities finish with a negative operating margin. Walk a median building through it: about $10.7 million of revenue produces roughly $980,000 of EBITDAR, capital and rent take close to $880,000, and what reaches the operating line is a rounding error. The 9% that looked like profit is mostly rent waiting to be paid.

EBITDAR is a coverage number, not take-home. What matters is how many times it covers rent and debt service, because the line underneath has no cushion.

For a leveraged buyer, that reframes the question. It is also why real estate and operations are so often split, opco and propco, with a REIT underwriting rent coverage while the operator underwrites census. A benchmark that stops at EBITDAR misses half the risk.

A negative operating margin is not proof a facility is failing. Nonprofit and government buildings, a quarter of the sector, run to a mission; opco/propco structures push profit into a related-party rent line; and single-year filings get revised. But the overlap of two weak signals, low occupancy and a sub-5% EBITDAR margin, is where genuine distress lives, and it is the supply that feeds turnaround plays and distressed M&A.

One layer explains a lot of the confusion around the sector. On Medicare fee-for-service patients alone, freestanding SNFs earn margins in the low-to-mid twenties by MedPAC's most recent read, while total all-payer margins sit near breakeven. Both are true: Medicare pays well above its own costs, and Medicaid and managed-care days pull the blend back to zero. Payer mix decides which reality a building lives in, which is why a strong Medicare census can carry a facility and a Medicaid-heavy one can sink it at the same occupancy.

None of this is a quirk of one dataset. The pattern matches the industry's leading annual cost-report study, which put 2024 median SNF operating margins just under 2% on median occupancy near 83%. This is a high-revenue, thin-and-volatile business where a good year and a bad one are a few points of census apart. Underwrite the facility, not the sector. Where any given building falls in the full distribution is what the Post-Acute tool exists to show.

Occupancy: the metric that moves the P&L

Check occupancy before anything else. Median SNF occupancy is 83.8%, recovered from pandemic lows near 78% and back above pre-COVID levels, and it tracks the margin more tightly than any other operating figure in the file. A SNF's cost base is largely fixed, so every empty bed comes straight off the margin.

Sort facilities into occupancy quartiles and the gradient is clean. The lowest quarter, below about 71%, runs a 5.8% EBITDAR margin and a negative operating margin near -3.5%. The top quarter, above roughly 92%, runs 12.0% EBITDAR and a positive 3.3% operating margin. Only the top quartile earns a clearly positive operating line; everything below it is fighting fixed costs.

That is why a buyer prices census risk first, and why the trend line usually matters more than any single year: a building at 82% and climbing is a different asset from one at 82% and sliding. One caution before reading the number too literally. Reported occupancy divides patient-days by certified beds, so an operator that has taken a wing offline can show a low figure while running its open beds nearly full. The reported rate is the right sector benchmark, but on a single target, reconcile it against available, staffed beds. Occupancy also swings by geography, a diligence angle in itself. For the state picture, see SNF occupancy by state; for how the figure is built, see how SNF occupancy is calculated.

Who operates the beds: chains, independents, and PE affiliation

Skilled nursing is less consolidated at the top than its reputation suggests. The file identifies 958 distinct operators, and no single one runs more than about 2.4% of the beds in it. The largest ten together account for 12.7% of facilities, the largest twenty-five for 21.2%. A long tail of regional operators runs the rest, which is exactly the fragmentation that makes the sector a roll-up target.

• Operator — Facilities

In-dataset facility counts, not each operator's full national footprint.

By in-file footprint, the largest operators are The Ensign Group (299 facilities), PACS Group (221), Life Care Centers of America (191), Genesis Healthcare (179), Creative Solutions in Healthcare (132), and Saber Healthcare Group (126). These counts reflect facilities present in this dataset, not each operator's full national footprint, so read them as an in-dataset ranking.

The chain-versus-independent line matters for a deal. A large operator brings management infrastructure, purchasing scale, and a playbook for turning around census, which is why chains can pay up for underperforming buildings an independent could not fix. The independents and small regional groups outside the top operators are the supply that feeds those roll-ups.

Private equity is present but hard to size from public data. The ownership flag captures identified sponsors only, so it undercounts real PE presence. Rather than publish a penetration rate the data cannot support, read PE through named operators: identified PE-backed groups in the file include Medilodge, Palm Garden Health and Rehabilitation, Mission Health Communities, and Aston Health. For the full operator picture, see the largest skilled-nursing operators.

Why the numbers vary (sanity checks, not formulas)

A handful of things move a facility off the national median, and none is a formula you can apply blind.

Size helps, up to a point. Median EBITDAR margin climbs from 3.0% at facilities under $5 million in revenue to 11.0% in the $10-to-$15 million band, then flattens near 11% above that. Bigger is not automatically more profitable, and revenue's correlation with margin is weak. What actually rises with size is occupancy, doing most of the work the revenue number appears to be doing.

Acuity and cost discipline separate top from bottom at any size. MedPAC's analysis finds higher-margin SNFs run more intensive-therapy days at lower cost per day, the same signal a high revenue-per-day carries. Clinical mix, not just how full a building is, sets how much each occupied day earns.

Payer mix sets the ceiling. A mostly-Medicaid building in a low-rate state is a structurally different operation from a Medicare-heavy one at identical occupancy. Medicaid covers roughly 62% of the sector's days on average, so state Medicaid rates are a first-order deal variable, and pending Medicaid funding changes are the risk the industry is watching most closely. Geography ties it together: rates and census swing enough by state that a facility three points below the national median can lead its market or trail it. Read a target against a local, same-sector comp, not the national line.

Using a benchmark responsibly

Cost-report data is the right tool for market position and the wrong tool for a valuation opinion. Keep the difference straight.

The figures are self-reported, unaudited, and lag current operations by 12 to 24 months, settling over time as filings are amended. Read them as directional market position, not audited accounting. They look backward, too: enrollment is shifting into Medicare Advantage, which pays SNFs below fee-for-service and compresses the blended rate faster than a two-year-old filing shows. Use medians rather than means, since facility-level margins swing wildly enough to make an average meaningless, and screen outliers before trusting a slice.

A benchmark tells you where a facility sits relative to its market. It does not tell you what the facility is worth. Getting from "40th percentile on occupancy, 60th on margin" to a defensible number is the work of a qualified appraiser, and it turns on facility-specific facts a benchmark cannot see. For how buyers make that move, see what a skilled-nursing facility is worth.

To place a specific facility or market against the full distribution, the Post-Acute tool returns median revenue, margin, and occupancy for any market on the free tier. The full percentile curves, per-facility financials, and the local comp set unlock on a plan.

Sources and method

• Source — CMS HCRIS provider cost reports, FY2024 and FY2025 filings
• Universe — 12,458 Medicare-certified SNFs across all 50 states
• Analysis set — 12,247 facility-level filings after dropping consolidated chain filings and margin-plausibility outliers
• Margin definition — EBITDAR is measured before rent; operating margin is after capital and rent
• Statistics — Medians and quartiles throughout; no means reported, since facility margins are too dispersed for a mean to carry meaning
• Payer mix — The ~62% Medicaid share is external industry context; SNF payer columns are empty in this extract
• Known limits — Self-reported, unaudited, and lagging current operations by 12 to 24 months; filings settle as they are amended

Full methodology and limitations

Disclaimer

This article is a market-data benchmarking resource derived from publicly available Medicare cost reports. It does not provide, and must not be relied upon as, a fair market value determination, valuation opinion, appraisal, or legal, tax, or compliance advice. Cost-report figures are self-reported, unaudited, and subject to revision. The opinion of value for any specific arrangement remains the professional judgment of a qualified appraiser.