Rental Vacancy Rate: What It Can—and Cannot—Tell You About Housing Supply

A rental vacancy rate measures the share of the rental inventory that is vacant and available for rent. In the U.S. Census Bureau's Housing Vacancy Survey, the national rental vacancy rate was 7.3% in the second quarter of 2026. The measure is useful for understanding how much rental housing is available relative to demand. It is also easy to misread.

Three distinctions matter from the start. First, the rental vacancy rate is not the share of all homes that are vacant. It is calculated against the rental inventory only, and many vacant homes are not for rent at all. Second, vacancy is not a measure of affordability: a market can have available units that many households cannot afford. Third, vacancy is not a forecast. A vacancy rate describes current availability, not where rents or prices will go next.

This article explains what the rental vacancy rate measures, how to read the latest Census data, why different Census vacancy figures should not be mixed, and what vacancy can and cannot tell you about housing supply.

About the data

This article draws on three different Census-based vacancy sources, which are not interchangeable. Current national and regional figures come from the U.S. Census Bureau's Current Population Survey/Housing Vacancy Survey (CPS/HVS), second quarter 2026. Longer-term context comes from the 2024 American Community Survey (ACS) 1-year estimates, as analyzed in the Census report Homeowner and Rental Vacancies in the American Community Survey: 2008–2024 (ACS-64). Metro Analytics local market pages use ACS 5-year estimates. Figures from one source are not compared with another as a trend.

What the rental vacancy rate actually measures

The Census Bureau's Housing Vacancy Survey defines the rental vacancy rate as the proportion of the rental inventory that is vacant for rent. The numerator is vacant year-round units available for rent. The denominator, the rental inventory, has three parts:

• renter-occupied housing units
• vacant year-round units rented but awaiting occupancy
• vacant year-round units for rent

Source: U.S. Census Bureau, CPS/HVS Second Quarter 2026 release, Explanatory Notes.

The denominator is what makes the measure useful. It compares available rentals with the rental market they belong to: units that are occupied by renters, already rented and waiting for a tenant to move in, or on the market now. It is not vacant rentals divided by all housing units. Owner-occupied homes, homes for sale, seasonal homes, and vacant units held off the market are all outside the rental inventory.

A hypothetical example shows how the pieces fit. Suppose a market's rental inventory contains 1,000 units: 900 occupied by renters, 30 rented but not yet occupied, and 70 vacant and for rent. The rental vacancy rate would be 70 divided by 1,000, or 7%. The market's owner-occupied homes and vacation properties would not change that figure at all, however many there are.

These numbers are illustrative only, not real data.

Two features of the definition are worth keeping in mind. The numerator counts only units that are vacant and available for rent, so a vacant unit that has already been leased is counted in the inventory but not as available. And the measure counts year-round units, so seasonal and vacation properties are excluded.

What the latest U.S. rental vacancy data shows

The Census Bureau estimated the national rental vacancy rate at 7.3% in the second quarter of 2026, compared with 7.0% in the second quarter of 2025. The Census Bureau did not find that difference statistically significant.

The homeowner vacancy rate was 1.2% in the second quarter of 2026, compared with 1.1% a year earlier. That difference was also not statistically significant.

Source: U.S. Census Bureau, CPS/HVS Second Quarter 2026 release, July 28, 2026.

The Census Bureau tests comparisons at the 90% confidence level.

The careful reading is that the survey does not show a confirmed change in either rate over the year. The point estimates are higher, but the Census Bureau's testing does not establish either difference as statistically significant. It would overstate the evidence to say national rental vacancy rose, or that rental markets loosened nationally, on the basis of these figures.

Two more cautions apply. HVS vacancy estimates are not seasonally adjusted, so small movements from one quarter to the next can reflect normal seasonal patterns and should not be read as a trend. And the two rates describe different inventories, a distinction covered below, so a rental vacancy rate of 7.3% and a homeowner vacancy rate of 1.2% are not directly comparable measures of the same thing.

Why vacant homes and rental vacancy are different statistics

A common misreading starts with the total number of vacant homes. In the second quarter of 2026, the Census Bureau estimated about 149.454 million housing units in the United States, of which about 15.643 million, or 10.5%, were vacant. That figure does not mean the rental vacancy rate is 10.5%. Most vacant homes are not for rent, and the two statistics use different denominators.

Total housing units149.454M(15.643M vacant (10.5%) of all housing units)

Vacant for rent

3.727M2.5%

For sale only

1.019M0.7%

Rented or sold, not yet occupied

1.005M0.7%

Held off market

6.466M4.3%

Seasonal

3.425M2.3%

Rental vacancy rate — share of rental inventory vacant for rent7.3%

This is a different question and a different denominator from the shares above — it divides vacant-for-rent units by the rental inventory only, not by all housing units.

Source: U.S. Census Bureau, CPS/HVS Second Quarter 2026, Table 3. Estimates are rounded and may not add to totals. Shares above are of all 149.454 million housing units, not of the rental inventory.

Only the "vacant for rent" category represents units that are empty and on the rental market, and it accounts for 2.5% of all housing units. The largest group, units held off the market, includes homes kept for occasional use, homes temporarily occupied by people whose usual residence is elsewhere, and units vacant for a variety of other reasons. Seasonal units are vacation and other part-year homes. Units for sale belong to the for-sale market. Units rented or sold but not yet occupied are already spoken for.

That 2.5% is not the rental vacancy rate either. It is the share of all housing units that are vacant for rent. The rental vacancy rate of 7.3% divides the same vacant-for-rent units by the much smaller rental inventory: renter-occupied units, units rented but awaiting occupancy, and units for rent.

The three figures answer three different questions:

• 10.5%: what share of all housing units are vacant for any reason?
• 2.5%: what share of all housing units are vacant and for rent?
• 7.3%: what share of the rental inventory is vacant and for rent?

For anyone assessing rental availability, the third question is usually the relevant one. The first can be useful for other purposes, such as understanding how much of the housing stock is seasonal or held off the market, but it is not a measure of rental supply.

The American Community Survey shows the same basic pattern in a different survey. The 2024 ACS estimated about 146.7 million housing units and 14.0 million vacant units. Of those, about 2.8 million were vacant for rent, about 849,400 were for sale, about 4.9 million were classified as "other vacant," and about 4.3 million were for seasonal, recreational, or occasional use. The Census Bureau notes that just over one-quarter of vacant units were available to rent or buy.

Source: U.S. Census Bureau, ACS-64.

These ACS figures are not directly comparable with the HVS counts above, for reasons explained below. The point they share is that a headline count of vacant homes is not a count of homes actually available on the rental or for-sale market.

Rental vacancy and homeowner vacancy answer different questions

The homeowner vacancy rate is built the same way as the rental rate, but for a different inventory. The Census Bureau defines it as the proportion of the homeowner inventory that is vacant for sale. The numerator is vacant year-round units for sale only. The denominator is owner-occupied units, plus vacant units sold but awaiting occupancy, plus vacant units for sale only.

Source: U.S. Census Bureau, CPS/HVS Second Quarter 2026 release, Explanatory Notes.

Rental vacancy vs. homeowner vacancy

Measure	What is counted as available	Inventory	Q2 2026
Rental vacancy rate	Vacant year-round units for rent	Renter-occupied units, units rented awaiting occupancy, and units for rent	7.3%
Homeowner vacancy rate	Vacant year-round units for sale only	Owner-occupied units, units sold awaiting occupancy, and units for sale only	1.2%

Swipe horizontally to view all columns →

Source: U.S. Census Bureau, CPS/HVS Second Quarter 2026 release, Explanatory Notes. Neither measure is presented as better or worse than the other.

Because the inventories differ, the rates should not be merged into a single "overall vacancy rate," and the gap between 7.3% and 1.2% should not be read as a direct comparison of how "tight" the two markets are. Each rate is meaningful relative to its own inventory and its own history.

Neither measure is more important in general. A reader studying rental housing, apartment operations, or renter household demand will usually focus on rental vacancy. A reader studying for-sale housing, homebuilding, or ownership demand will usually look at homeowner vacancy alongside other for-sale indicators. Many questions call for both, read separately.

Also note that the homeowner vacancy rate counts only vacant units for sale. It does not capture homes listed for sale while their owners still live in them, a limitation discussed below.

Why Census vacancy rates can differ across surveys

The Census Bureau publishes vacancy rates from more than one survey, and the numbers are not the same. That is expected, not an error, and it is one of the most common sources of confusion in housing analysis.

The two sources most readers encounter are the quarterly Current Population Survey/Housing Vacancy Survey (CPS/HVS), which produced the Q2 2026 figures above, and the annual American Community Survey (ACS).

The Census Bureau's 2026 report on ACS vacancies, Homeowner and Rental Vacancies in the American Community Survey: 2008–2024 (ACS-64), reports these 2024 ACS rates:

ACS vs. CPS/HVS vacancy estimates

Survey and period	Rental vacancy	Homeowner vacancy
2024 ACS 1-year estimates	5.7%	1%
CPS/HVS, Q2 2026	7.3%	1.2%

Swipe horizontally to view all columns →

Sources: U.S. Census Bureau, ACS-64 (May 27, 2026); CPS/HVS Second Quarter 2026 (July 28, 2026). These rows are not a time series and must not be read as a change from 2024 to 2026.

It would be wrong to conclude from this table that rental vacancy "rose from 5.7% to 7.3%." The two figures come from different surveys, cover different periods, and are produced in different ways. Any change over time has to be measured within one survey.

The Census Bureau's guidance on vacancy data sets out the main differences between the surveys:

• Purpose. The HVS exists largely to provide timely quarterly estimates of vacancy and homeownership rates. The ACS provides annual estimates of a wide range of social, economic, and housing characteristics for many geographic areas.
• Geography. The HVS publishes estimates for the nation, regions, states, and large metropolitan areas. The ACS produces estimates for most census geographies, down to small areas, with estimates for smaller places based on multiyear data.
• Sample and collection. The surveys use different samples and interview methods, and the ACS is mandatory while the CPS/HVS is voluntary.
• Timing. The HVS records a unit's vacancy status as of the interview week each month. The ACS classifies units as of the day they are first contacted, and most vacant units in the ACS are not identified until a later stage of data collection.
• Residency rules. The ACS counts a unit occupied entirely by people staying two months or less as vacant, with a current residence elsewhere. The HVS applies a usual-residence rule.

Source: U.S. Census Bureau, Vacancy Rate Fact Sheet.

The practical rule is simple: use each survey consistently within its own context. The HVS is suited to current national and regional conditions and to tracking change quarter by quarter or year over year within the HVS series. The ACS is suited to detailed geographic comparisons and longer-term local context. Neither is "the correct" vacancy rate. They are different measurements designed for different uses.

What a low or high vacancy rate can tell you

The Census Bureau describes vacancy rates as indicators of housing availability relative to demand and notes that some vacancy is necessary in a functioning housing market because renters and buyers need units available for occupancy.

A very low vacancy rate can therefore be consistent with limited immediately available inventory, but vacancy does not capture every unit being marketed.

Within that framing, vacancy supports some cautious readings:

• A relatively low rental vacancy rate can be consistent with limited available rental inventory relative to demand.
• A relatively high rental vacancy rate can be consistent with more available rental inventory relative to demand.

The key words are "relatively" and "can be consistent with."

"Relatively" means relative to something specific: the same market's own history, comparable markets measured with the same survey, or a comparable segment of the housing stock.

"Can be consistent with" means the vacancy rate is compatible with that reading but does not prove it.

A higher rate might reflect a wave of recently completed buildings that have not yet leased up, a fall in household demand, a mismatch between the units available and what renters can afford, or a mix of these.

A lower rate might reflect strong demand, limited construction, or a stock of units that rarely turn over.

The vacancy rate alone cannot tell you which explanation applies.

Why there is no single "good" vacancy rate

Searches for vacancy rates often lead to a single number presented as "healthy" or "ideal," with ranges above or below it labeled too tight or too loose.

The Census sources used in this article do not establish any such universal threshold, and this article does not offer one.

There is no single vacancy rate that automatically defines a healthy housing market.

The reason is that the same rate can mean different things in different settings.

Interpretation depends on context, including:

• Location and geography: a national average, a region, a metro area, and a neighborhood can behave very differently.
• Rental stock and property type: single-family rentals, small multifamily buildings, and large apartment complexes can have different typical levels of turnover and vacancy.
• Market structure: the mix of owners, unit sizes, and price points in a market shapes how units are marketed and filled.
• Historical norms: a rate that is unusual for one market may be typical for another. A market's own history, measured consistently, is often a more useful reference than a generic benchmark.
• Recent supply additions: newly completed units that are vacant and being offered for rent can contribute to measured vacancy while they are being leased.
• Household demand: population change, household formation, and migration affect how many households are looking for rental housing.
• Affordability and income: available units may not match what local renters can pay.
• Tenure patterns and local economic conditions: the balance between renting and owning, and changes in employment and incomes, shape rental demand.

A useful reading of any vacancy rate starts with a comparison: to what, measured how, and over what period?

A number without that context cannot be sorted into "good" or "bad."

Vacancy differs across regions

National rental vacancy combines very different regional estimates.

In the second quarter of 2026, the Census Bureau reported:

South

9.5%

Midwest

6.9%

Northeast

5.9%

West

5.3%

Northeast and West estimates were not statistically different from each other. Source: U.S. Census Bureau, CPS/HVS Second Quarter 2026, Table 2.

Although the Northeast and West estimates differ by 0.6 percentage points, the Census Bureau found them not statistically different from each other, so the data do not establish which of the two had the lower rate. The Census Bureau reports the South as having the highest rate, followed by the Midwest.

Compared with the second quarter of 2025, the Census Bureau found the Northeast's rental vacancy rate higher, while the Midwest, South, and West rates were not statistically different from a year earlier.

The release also reports rental vacancy by location within and outside metro areas:

None of these rates was statistically different from the second quarter of 2025.

Source: U.S. Census Bureau, CPS/HVS Second Quarter 2026.

These are descriptive estimates, not judgments about market quality. A higher regional rate does not by itself show that a region is oversupplied, and a lower rate does not show that it is undersupplied or that its housing market is stronger. Regions differ in their housing stock, rental share, pace of construction, population change, and incomes. A regional rate is also an average across very different metros and neighborhoods, so a single regional figure can conceal wide local variation. Explaining why the regional rates differ would require evidence this survey does not provide.

What vacancy cannot tell you on its own

Vacancy is a snapshot of availability. On its own, it cannot establish:

• whether rents will rise or fall
• whether home prices will increase or decrease
• whether housing is affordable to local households
• whether a market is overbuilt or undersupplied
• whether new construction should slow or accelerate
• whether a particular investment will perform well
• how housing demand or housing values will change in the future

Each of these questions needs additional evidence: rent and price data, income and affordability measures, construction and permitting activity, household formation, employment, and financing conditions. A vacancy rate can be one input to those judgments, but it cannot settle any of them, and it carries no forecast.

Vacancy is also an incomplete measure of supply. By definition, it counts units that are empty. But not all available housing is vacant. The Census Bureau notes that occupied homes can also be part of the available supply. A home can be listed for sale while its owners still live in it. A rental unit can be marketed to new tenants before the current tenant moves out.

Source: U.S. Census Bureau, ACS-64.

Those units are on the market but do not appear in a vacancy statistic. The result is that vacancy captures an important but partial view of housing availability. It measures one slice of the market—units that are empty and on offer—and leaves out occupied units that are also for rent or sale. For that reason, vacancy is best read alongside other supply measures rather than as a complete count of what is available.

Rental vacancy

+

New supply

+

Household demand

+

Affordability

+

Income & financing

↓

Broader housing-market context

Conceptual illustration only — not a formula, score, or weighting of Metro Analytics' Market Score.

How Metro Analytics uses vacancy

Metro Analytics treats rental vacancy as one market signal, not as a verdict on a market.

In the current Metro Analytics Market Score methodology, the Supply & Demand component uses the Census ACS rental vacancy rate as a proxy for inventory pressure. That component carries 25% of the overall Market Score.

The 25% weight belongs to the Supply & Demand component as defined in the methodology, not to the vacancy rate as a standalone figure, and vacancy is not the entire Supply & Demand assessment.

Source: Metro Analytics Methodology.

The local vacancy values on Metro Analytics market pages come from ACS 5-year estimates. These pool several years of survey responses, which makes them suitable as a longer-term signal of local availability and inventory pressure, particularly for smaller areas. They are a different product from the quarterly CPS/HVS figures used for national and regional conditions in this article. The two differ in survey, estimate period, and geography, so a local ACS 5-year vacancy rate on a market page should not be compared directly with the national Q2 2026 HVS rate of 7.3%.

Market Scores also describe current conditions. They are not forecasts. This article explains how to interpret vacancy. For the conditions in a specific market, a market page is the better place to look, because it shows vacancy alongside the other measures that give it context.

The takeaway

Rental vacancy is useful because it shows how much of the rental inventory is available at a point in time, relative to the rental market it belongs to. But it is one signal, not a complete picture.

Reading it well means knowing which dataset produced it, what geography it covers, and what type of vacancy it measures. It also means placing it beside other evidence: available supply beyond vacant units, recent construction, household demand, income and affordability, market structure, and financing and economic conditions.

Rental vacancy is a useful indicator of housing availability relative to demand, but it only becomes meaningful when readers understand the dataset, the denominator, the geography, and the broader market context.

To see vacancy alongside those signals for a specific area, you can explore Metro Analytics Markets, or read the Methodology to see how vacancy fits into the Supply & Demand assessment.