Rental Property Mortgages: 37.5% Have One, Census 2024 Survey

What the survey counts, and how we read it

The Rental Housing Finance Survey (RHFS) is, in HUD's words, “the most comprehensive survey of rental housing properties in the United States”. The Census Bureau runs it for HUD every three years. Its methodology defines a property as “one or more buildings owned by a single entity (person, group, leasing company, and so on)”, so the unit of count is the property as the survey defines it, not the loan and not the landlord. The 2024 survey asked about calendar year 2023, collected answers from June to November 2024, and had an average unit response rate of 45.1%. Census released the data on February 19, 2026. The public use file holds 4,425 property records, of which 825 are one-to-four-unit properties; each carries a weight that scales it to the 18,965 thousand rental properties in the country, plus 100 replicate weights for margins of error.

Everything below that is not marked “published” is our tabulation of that file with the survey weight, and every margin of error is ours, 90% confidence, from the replicate weights with Census's method (successive difference replication, 4 divided by 100 times the sum of squared differences, times 1.645). We checked the method against Census: our weighted counts and margins reproduce its published key estimates exactly (6,502 thousand properties with a mortgage, margin 943; 10,814 thousand without, 531; 1,649 thousand not reported, 643; 18,965 thousand in all, 271), and our debt service coverage medians reproduce the five figures in its infographic. Census's March 2026 correction to the occupancy and vacancy tables does not touch the mortgage variables used here, and the file still carries version 1.0.

How many rental properties carry a mortgage

The question looks simple and has two answers, because 8.7% of properties (plus or minus 3.4 points) did not say whether they have debt. Counting only those that answered, 37.5% do. Counting every property and treating a missing answer as no, 34.3% do. Neither is a loan count: a property with a first mortgage and a home equity line is one property.

Rental properties with a mortgage or similar debt, by size of property, 2024 survey (calendar year 2023)

Margins are 90% (our arithmetic). The gradient is steady, and from 25 units up about two-thirds of the properties that answered have debt. The step from one unit to two-to-four units (34.8% to 50.3%) is larger than the margin on either estimate. The gap between the 1-4 unit share (37.1%) and the national share (37.5%) shows how little the 485 thousand properties of 5 units or more move the total, because one-unit properties are 15,714 thousand of the 18,965 thousand. Most of the 6,502 thousand mortgaged properties are houses: 5,016 thousand one-unit properties (plus or minus 922 thousand).

What kind of debt it is: of properties with debt whose first-debt type was reported, 94.9% (plus or minus 3.2) have a mortgage, deed of trust or trust deed, 2.5% (plus or minus 2.1) a home equity line or loan secured by the rental, 2.5% (plus or minus 1.7) some other secured loan and 0.1% a contract to purchase or land contract. Two or more debts are rare on small properties, 2.5% of mortgaged 1-4 unit properties (plus or minus 1.4), and common on large ones, 12.1% at 5 or more units (plus or minus 4.9).

HUD's release for the 2021 survey said that “About 41 percent of all rental properties have a mortgage or similar debt”, against 35 percent in 2018. We do not read our 37.5% (or 34.3%) as a fall from 41%: the release does not say how it treated properties with no answer, the 2024 file leaves 8.7% unanswered, and the interviews covered different years. What the three surveys agree on is the order of magnitude: between a third and two-fifths of rental properties carry debt.

Who the first lender is

For each property with debt the survey asks where the first mortgage was obtained and lets the owner pick one of ten lender types or “don't know”. Lender type is missing for 11.8% of mortgaged properties (550 of the 2,162 mortgaged records, weighted), so the shares below are of properties whose type was reported, which includes the 2.3% who answered don't know.

Lender type of the first debt, share of mortgaged properties that reported one, 2024 survey

Margins are 90% plus or minus points (our arithmetic). Three readings the data supports. First, two lender types account for most of it: commercial banks (44.6%) and mortgage banks or mortgage companies (26.6%), together 71.2%. Second, size changes the mix: for 1-4 unit properties the mortgage company share is 27.2% (plus or minus 7.1), for 5+ units 8.5% (plus or minus 2.7), where commercial banks hold 66.1% (plus or minus 3.6) and state, municipal and housing-finance agencies another 3.8% (1.9% each, our arithmetic). Third, below the top two the ranking is not reliable: savings institutions and credit unions both come out at 9.7% for all sizes, but with margins of 10.1 and 7.0 points, and finance companies at 3.9% (plus or minus 4.4).

What the table does not say matters as much. The survey records a lender type, not a lender and not a loan program: a rental-property loan from a non-bank investor lender would be classed as a mortgage company, a finance company or “other” depending on how the owner answered (our reading), so nothing here sizes the DSCR or hard-money market; for named lenders see our DSCR lenders ranked by 2025 loans. Seller financing is nearly absent as a first mortgage: individuals are 0.4% of first lenders and a contract to purchase 0.1% of first-debt types (see our seller financing page for the legal side).

Why owners with no mortgage have none

Of the properties reported as having no mortgage or debt, the survey asks why. Three answers are offered, and the codebook words the first two as “I prefer not to carry any debt on this property or do not want a mortgage” and “I have tried to get a mortgage, but was denied”; the third is “other”, with the specified text not in the public file.

Why properties with no mortgage have none, share of properties that gave a reason, 2024 survey

Margins are 90% (our arithmetic). The denial answer was given by 9 of 1,457 records and all 9 are in properties of 5 or more units; weighted, that is about 500 properties nationwide (plus or minus 500) and about 0.005% of unmortgaged properties. Properties with no debt whose owners say they prefer none number about 8.0 million (plus or minus 1.1 million), 42.0% of all rental properties (plus or minus 5.8 points, our arithmetic).

Read carefully, the figure says that among today's owners, being refused a mortgage is not what keeps a rental free of debt. It does not say that credit is easy: the survey interviews current owners, so a would-be buyer who was refused a loan and did not buy is not in it (our reading), and the 26.0% “other” group (plus or minus 9.8) could hold reasons such as an inheritance, a gift or a property bought for cash that nobody needed to finance. The practical reading for a small landlord is that going without a mortgage is an ordinary, deliberate choice, made by roughly three of every four owners who have none.

Cash down payments and recent buyers

The survey asks whether a cash down payment was made, but only for properties acquired in the last six years: the codebook comment reads “Only applies to the last six years”, and the universe is properties acquired in 2019 to 2024. The answer is yes or no, and 8.0% of those buyers (plus or minus 5.1 points) left it blank.

Cash down payment, properties acquired 2019 to 2024, 2024 survey

The estimate is imprecise: 470 records answered, and the margin on the all-size share is 14.5 points, so anything from about 34% to 63% is consistent with the survey. The dollar amount is just as loose. What the file does not hold is a down-payment percentage or a loan-to-value ratio, and it does not say what the “no” answer means in practice (our reading: a purchase or transfer with no cash at closing, but the codebook does not say). For percentages from real loans, our investment property down payment study uses the federal mortgage record: a median combined loan-to-value of 75% on 326,533 purchase loans in 2025.

How much of the mortgage payment the rent covered

The public file includes a debt service coverage ratio that Census computes for each mortgaged property: rent collected plus commercial rent, minus operating expenses (taxes, insurance, utilities, repairs, management and the rest), divided by the regular mortgage payments. A ratio of 1.00 means net operating income equals the payments. The ratio could be computed for only 817 of the 2,162 mortgaged records (201 of 324 at 1-4 units and 616 of 1,838 at 5 or more), because it needs rent, expenses and payment answers together.

Median debt service coverage of mortgaged rental properties, calendar year 2023 (ratio, our tabulation)

For small properties the median is under 1.00: 0.80 for 1-4 units (plus or minus 0.25), and 55.5% of them (plus or minus 12.6 points) had net operating income below their mortgage payments in 2023. Large properties covered their payments with room to spare, at a median of 1.64 for 50 units or more (plus or minus 0.26). Two cautions before using this as a benchmark. The ratio is not a DSCR lender's ratio: lenders define theirs their own way (see our DSCR loan requirements), and the file has fields showing whether the regular payment includes property taxes and insurance, which would put those costs on both sides of Census's ratio (our reading). And the figure is before depreciation, income tax and capital spending. The finding that fits the data is narrower than a claim that small landlords lose money: the typical mortgaged small rental in 2023 did not cover its payment from net operating income as the survey measures it, with a margin that is wide on 201 records.

When the mortgage was placed, and who holds it

For mortgaged 1-4 unit properties the file reports the year the first mortgage was placed, in bands. A third of them, 33.6% (plus or minus 7.0), were placed in 2019 to 2021, and 11.3% (plus or minus 4.8) in 2022 to 2024, so 44.9% date from 2019 onward (our arithmetic). The year can be a refinance rather than a purchase, and the file has no interest rate, so it says when the debt was taken, not at what price.

Year the first mortgage was placed, mortgaged 1-4 unit properties, 2024 survey

Ownership type changes the picture only for larger properties. Among 1-4 unit properties the mortgage share is 39.7% for individual investors (plus or minus 10.2) and 27.0% for LLPs, LPs and LLCs (plus or minus 14.8); the margins overlap, so the file cannot show a difference. At 5 or more units, 66.3% of LLP, LP or LLC properties have a mortgage (plus or minus 4.4) against 40.1% of those owned by individuals (plus or minus 7.3). For how the entity choice interacts with lenders, see our LLC for rental property study, which tabulates who owns the same properties from the same file. For 5+ unit properties, about half of first loans (49.6%, plus or minus 4.6) were originally between $1 million and $25 million, the only loan-size field the public file offers, and only for that size class.

What a reader can do with this

1. Use the right base when you quote a share. “About a third of rental properties have a mortgage” (34.3% of all) and “37.5% of those that answered” are both defensible; “a majority have none” is also true. Say which one, and for small landlords use the 1-4 unit figure (37.1% of those that answered, plus or minus 4.4).
2. Treat a cash purchase as the common case for small rentals, not the exception. 62.9% of reporting 1-4 unit properties have no debt (our arithmetic), and three-quarters of the owners with none say it is by preference. If you plan to buy for cash, you are in the majority; if you plan to borrow, so are the 37%.
3. Expect to shop between a bank and a mortgage company. Those two types are the first-mortgage lender on 71.2% of mortgaged properties, and for houses and small multiples the mortgage company is a quarter of the market. A DSCR or investor-program lender may sit in either column, so compare loan terms rather than lender labels; our conventional versus DSCR comparison and investment property loans guide set out the choices.
4. If you own a rental free and clear, you hold an option, not a gap. A property with no debt can be borrowed against later; our cash-out refinance page covers the rules and costs. Whether to do it depends on your rate, cash flow and risk tolerance, not on what the average landlord does.
5. Compute your own coverage before you borrow. The survey's small-property median under 1.00 is a reminder that a payment can exceed net rent on a thinly financed deal; divide your expected net operating income by the full annual payment and compare it with the lender's minimum.

If you need a loan for a one-to-four-unit rental, the lenders below make investor loans on rentals; ask each what it lends on. Kiavi pays us a referral fee when a loan closes through its button in the box below. This is analysis of public documents, not investment, legal, tax or lending advice, and not a loan offer.

What the survey cannot tell you

• It counts properties, not loans or landlords. The survey's definition of a property (buildings owned by a single entity) decides what counts as one, and a property with two debts appears once.
• Mortgage status is missing for 8.7% of properties (plus or minus 3.4) and 20.9% of the largest class, and lender type for 11.8% of mortgaged properties. We left unanswered records out of each base and say so in the CSV.
• It is one year, self-reported. The reference year is 2023; the unit response rate was 45.1%; Census adjusts the weights for nonresponse, but answers about debt are the owner's.
• The public file has no interest rate, loan-to-value ratio, credit score or loan program, and recodes some fields: the first-loan amount is in nine bands and reported only for properties of 5 or more units.
• Margins are wide for subgroups. The file holds 825 one-to-four-unit records, 324 of them mortgaged, and fewer for each lender type; where a margin is as large as the estimate, we do not rank.

FAQ

Sources: U.S. Census Bureau and HUD, Rental Housing Finance Survey 2024 public use file (rhfspuf2024.csv, version 1.0, February 19, 2026), Codebook for 2024 Internal and Public Use Files (version 1.0, February 2026), RHFS 2024 methodology page and full methodology (Table 5 key estimates, definitions, reference period, response rate, variance method), 2024 RHFS infographic, PUF versioning document and the Census news and updates notice of March 19, 2026; HUD news release 22-242 of November 29, 2022 (2021 survey). Weighted shares, margins of error (90%, successive difference replication with the file's 100 replicate weights), medians and the combinations of categories are our arithmetic, with the script saved. Everything was read on October 29, 2026. This is analysis of public documents, not investment, legal, tax or lending advice.