Units vs Houses in Australia 2026: Why Units Are Quietly Outperforming — and When a House Is Still the Better Buy

Published: 1 August 2026

At a Glance: Units vs Houses, June 2026

Source: PropTrack Home Price Index, June 2026 (national medians and annual growth); Cotality June 2026 rent and yield data; our analysis. Yield premium is indicative — verify per suburb.

Somewhere in the past eighteen months, the oldest rule of thumb in Australian property — buy the house, land appreciates, buildings depreciate — quietly stopped describing what the market was actually doing.

On PropTrack's June 2026 index, national unit values are up 6.7% over the year against 5.6% for houses. Through the three months of price falls since the March peak, units have consistently recorded smaller declines than houses. Brisbane's lower-quartile units — the cheapest quarter of the apartment market — have become, on Cotality's June Housing Chart Pack, the most expensive entry-level apartments in the country, overtaking Sydney's. And in Melbourne, unit yields have climbed from near the bottom of the capital-city table two years ago to third-highest today.

None of this happened because Australians suddenly fell in love with apartments. It happened because three separate forces — an affordability ceiling, a yield repricing and a supply squeeze — all pushed demand toward the same stock at the same time, right as tax reform handed new builds (mostly units) a structural advantage. This article walks through each force with the data, then does the harder job: separating the units that deserve the momentum from the ones that will still disappoint, and setting out when a house remains the unambiguous answer.

If you're earlier in the journey, our beginner's guide to property investment and where should I buy in 2026 cover the fundamentals this builds on.

What the Data Actually Shows in 2026

Start with the national picture. PropTrack's June 2026 release has national home prices down 0.3% for the month — the third consecutive fall — but the composition tells the real story: the declines are concentrated in established houses in Sydney and Melbourne, while units nationally posted smaller falls and stronger annual growth (+6.7% vs +5.6%). Cotality's June index shows the same downturn shape (−0.4% nationally), and its rental series has national rents up 5.9% for the year against vacancy near record lows.

City-level splits sharpen it further. Earlier in 2026, PropTrack-attributed reporting had capital-city units up 8.4% over the year against 7.5% for houses; Brisbane units at one point were running above +20% year-on-year against mid-teens for Brisbane houses, and Perth units were outpacing even Perth's exceptional house growth. Those specific prints have cooled with the broader market since — but the pattern (units growing faster than houses within the same city) has persisted through the peak and into the downturn.

Two details in the data get missed:

Units are behaving defensively, not just growing faster. In a downturn led by rate-driven borrowing-capacity compression, the cheaper asset class has the deeper buyer pool underneath it. When a house buyer's capacity falls 10–12%, they don't leave the market — they step down a price tier. The unit market is where those buyers land, which cushions unit prices at precisely the moment house prices are being marked down.

The entry-level end is the hottest end. Cotality's June Chart Pack finding that Brisbane's lower-quartile units now out-price Sydney's as the country's most expensive entry-level apartments — with barely $2,000 separating the income needed for a median Brisbane unit versus a median Sydney one — is a snapshot of demand crowding into the cheapest purchasable stock in the growth capitals. That is affordability-ceiling behaviour, textbook and measurable.

For the monthly numbers as they update, our PropTrack HPI tracker and Home Value Index tracker carry the latest splits with every release.

Driver One: The Affordability Ceiling Is Doing the Buying

Run the arithmetic that every mortgage broker in the country is running. The February–May 2026 rate hikes took the cash rate from 3.60% to 4.35%; on our modelling that stripped roughly 10–12% from typical borrowing capacity (mechanics in our borrowing capacity guide). A dual-income household that could finance ~$950,000 of purchase in late 2025 now tops out somewhere in the mid-$800,000s. The median capital-city house costs over a million dollars. The median unit costs $735,000.

The buyer didn't choose the unit over the house. The serviceability calculator chose it for them.

This mechanism has three properties that make it durable rather than a passing quirk:

1. It's ratcheted by policy. Expanded 5%-deposit schemes for first home buyers (our analysis here) operate under price caps in each city — caps that sit at or below unit medians and far below house medians in Sydney, Brisbane and Melbourne. Subsidised demand is steered toward units by design.
2. It doesn't unwind until rates fall meaningfully. The June quarter CPI has all but confirmed the cash rate has peaked at 4.35% — but with underlying inflation at 3.6%, no major bank forecasts cuts before 2027 (our CPI analysis covers why). The capacity squeeze that redirects demand to units is a 2026 and 2027 story.
3. It compounds with investor behaviour. Yield-driven investors — the subject of the next section — hunt the same sub-$800,000, high-yield stock the capacity-constrained owner-occupiers are bidding on. Two distinct buyer types, one price bracket.

Driver Two: The Yield Case Has Never Been Stronger This Cycle

On Cotality's June 2026 data, the national gross yield has lifted off its 3.55% cycle low to 3.59% and is expanding — the first cyclical yield expansion since 2023, driven by the rare combination of falling values and rents still compounding at 5.9% against 1.3% national vacancy (SQM June data).

Within that, the unit premium is structural. A worked comparison at June 2026 medians:

Source: medians per PropTrack June 2026; yields indicative of typical capital-city house/unit spreads per Cotality data — verify per suburb before relying on them. Interest at the ~6.4% average investor rate; excludes strata, rates, insurance, management and tax effects.

The pattern holds even after you add back strata levies (typically $3,000–$8,000+ a year, more in facility-heavy towers): the unit's smaller debt and higher yield leave a materially shallower annual hole, funded from a deposit roughly $66,000 lighter. In a market where Cotality counts fewer than 1% of suburbs as positive-cash-flow, "least negative" is the operative competition — our positive cash flow guide works the full after-tax version.

Three sharpeners on the yield case:

• Rent growth is skewing toward units too. The same affordability pressure operating on buyers operates on tenants: as house rents outrun budgets, tenant demand steps down into units, and unit rents in the tight inner and middle rings have been re-accelerating. Vacancy risk is not uniform, though — suburban and boutique-block units let fast at 1.3% national vacancy, while CBD investor towers remain the segment where vacancy spikes first in any downturn, as the pandemic years demonstrated.
• Melbourne is the live case study. Cotality notes Melbourne unit yields have moved from second-lowest among the capitals two years ago to third-highest today — a repricing produced by years of flat unit values against relentless rent growth. Yield-led buyers are now arriving; our Melbourne recovery guide makes the counter-cyclical case.
• Serviceability likes yield. Lenders credit rental income (typically shaded to ~80%) in assessment. Higher-yield stock partially funds its own borrowing capacity — one reason the same investor can often finance a unit purchase they couldn't finance as a house.

Driver Three: The Supply Squeeze Protects Unit Incumbents

The historical knock on units is that supply is elastic: whenever unit prices rise, developers manufacture more of them, capping growth — the mechanism that buried the 2016–19 investor cohort in Brisbane and Melbourne under a wave of new towers.

That mechanism is jammed. Apartment approvals fell ~30% in original terms in May 2026, and the reason isn't demand — it's feasibility. New-dwelling construction costs rose 5.8% in the year to June (ABS, and still accelerating), while sale prices nationally are falling. A developer pricing a 2027 completion is squeezed from both ends, and marginal projects are being shelved rather than started. Presales requirements, construction-sector insolvencies and financing costs at a 4.35% cash rate do the rest.

For an owner of existing unit stock, this is as favourable as the supply picture has looked in a decade and a half:

• The 2027-28 competing pipeline is thin — in rentals (supporting the 5.9% rent trajectory) and in resale (fewer shiny new competitors at sale time).
• Replacement cost keeps rising under existing values. When building the equivalent unit costs 5.8% more each year, established units get repriced upward by reference — the scarcity-premium effect that usually protects houses now working for attached stock.
• The squeeze is self-reinforcing near-term. The cost line that would need to fall to restart the pipeline is the same line the CPI shows accelerating. This does eventually resolve — government supply incentives and an eventual rate-cut cycle will restart approvals, with a ~2-year construction lag. The protected window is roughly 2026–2029, not forever, and approvals data is the early-warning indicator to watch.

We've Seen Unit Booms Before — Why This One Is Built Differently

Any recommendation of units in Australia has to answer for the last cycle, so here is the comparison laid out honestly. And the pattern is older than one cycle: units have swung in and out of favour repeatedly — outperforming briefly in the post-GFC first home buyer stimulus years, lagging badly through the 2010s supply wave, then diverging sharply through COVID as buyers paid up for space and houses left apartments behind. Unit outperformance is cyclical, not unprecedented — what varies is how long each phase runs.

Through the mid-2010s, off-the-plan apartments were the country's most heavily marketed investment product. Developers responded to the demand exactly as economics predicts: apartment completions hit record highs in 2016–18, concentrated in Brisbane's inner ring, Melbourne's CBD and Docklands, and Sydney's growth corridors. The result was measured and brutal — inner-Brisbane and Melbourne-CBD unit values went sideways or backwards for the better part of five years, rents stagnated under vacancy spikes, and settlement valuations routinely came in under contract prices. Investors who bought the marketing rather than the suburb waited nearly a decade to break even. That experience is why "never buy a unit" hardened into pseudo-wisdom for a generation of investors.

Compare the setup mechanically:

Source: our analysis; ABS building approvals and CPI data; SQM Research vacancy series; Cotality rent series.

The last row is the one to keep. The 2016–19 wreck was not evenly distributed: boutique, land-rich, owner-occupier-grade units in landlocked suburbs sailed through it, because they never competed with the cranes. The stock that failed was the stock the current filter (small schemes, land share, owner-occupier majority) is designed to exclude. History's lesson isn't that units fail — it's that substitutable units fail whenever supply arrives, and the 2026 twist is simply that supply, for once, cannot arrive quickly.

Financing a Unit vs a House in 2026: What Lenders Actually Do Differently

The serviceability side favours the unit, as the worked ledger above showed: smaller debt, and rental income (which lenders credit at ~80% in assessment) covering a larger share of the repayment. But the collateral side is where unit buyers hit rules house buyers never see:

• Minimum size. Most lenders draw a hard line around 40–50m² of internal living area (excluding balconies and car space). Studio and micro-apartments — often the highest-yielding stock on paper — are unfinanceable at standard LVRs with most of the majors, which caps their resale market to cash buyers and specialist lenders. This alone explains much of their yield premium: it's an illiquidity payment, not free income.
• Density and postcode limits. Lenders maintain internal lists of high-density postcodes and specific buildings where exposure is capped. In a flagged postcode, the same borrower who qualifies for 90% LVR on a villa may be limited to 70–80% on a tower apartment — turning a $147,000 deposit into a $220,000 one on a $735,000 purchase.
• New-building caution. Some lenders restrict lending in buildings under a certain age or where one developer/investor holds a large share of the units. Off-the-plan purchases add valuation risk at settlement: if the bank's valuer marks the unit below contract price, the shortfall comes from your deposit.
• Strata levies count against you in servicing. Lenders deduct ongoing strata levies as a committed expense in serviceability assessment — a $6,000-a-year levy reduces borrowing capacity in a way a freestanding house's (self-managed, deferrable) maintenance never does. High-levy buildings are penalised twice: in your cash flow and in your capacity.
• Where the standard rules still apply. APRA's 3-percentage-point serviceability buffer, the high-DTI limits (our APRA DTI guide covers the 20% cap on ≥6× debt-to-income lending), lenders mortgage insurance above 80% LVR, and investor-vs-owner-occupier rate premiums (typically 30–60 basis points) apply to units and houses alike. Loan structure choices — interest-only vs P&I, fixed vs variable — are likewise property-type-neutral.

The practical sequence: get the lender question answered for the specific building type before falling in love with the property. A broker query costs nothing; discovering a postcode restriction after exchange costs the deal. And note the alignment between what banks will lend against and what this article's selection filter recommends — small schemes, larger formats, owner-occupier buildings finance like houses precisely because lenders have priced the same risks this article describes. When the credit department and the investment thesis agree, that is usually signal.

Deposit mechanics, LMI thresholds and the guarantor routes are covered in our deposit guide; if the deposit itself is the constraint, equity-release from an existing property is the standard investor path.

The Tax-Reform Angle: Reform Quietly Favours (New) Units

The 2026 reform package — legislated in June, effective 1 July 2027 — redrew the after-tax comparison between property types in a way most commentary hasn't priced. The full mechanics are in our negative gearing and CGT changes guide; the piece that matters here:

• Buy an established house or unit after 12 May 2026 → from 1 July 2027 its rental losses quarantine against other property income (carrying forward otherwise), rather than offsetting salary.
• Buy a qualifying new build → negative gearing continues as before. And the new-build pipeline is dominated by attached stock: apartments, townhouses, house-and-land in fringe estates.

Combine that with depreciation and the after-tax gap widens further: a new unit carries full Division 43 capital works deductions plus Division 40 plant-and-equipment on brand-new assets (established dwellings lost second-hand plant deductions back in 2017 — see our depreciation guide). A new $735,000 unit can generate $10,000–$15,000+ of first-year non-cash deductions and keep gearing against salary post-2027; an established $1,001,000 house bought today gets neither.

Our analysis: this doesn't make new units automatically superior — developers price incentives into new stock, and the premium you pay for "new" can exceed the tax benefit (we model exactly this trade-off in our new-build vs established analysis). But at the margin, the reform pushes every post-2027 negatively-geared investment dollar toward the new-build market — which is a unit market. Expect that flow to show up in entry-level unit demand from 2027.

The rest of the tax ledger has genuine property-type differences worth pricing:

• Land tax quietly favours units. State land tax assesses the land component, and a unit's share of land value is a fraction of a house's — so a portfolio of units can sit under a state's land tax threshold where a single house in the same suburb exceeds it. Over a multi-property holding, that is thousands a year of difference; our land tax guide maps the state thresholds.
• Strata levies are deductible; so are capital works. The administrative and sinking-fund components of levies are generally deductible in the year paid (special levies for capital improvements instead depreciate as capital works), and units in buildings constructed after 1987 carry Division 43 deductions on the build cost. The full deductions list covers what landlords routinely miss.
• CGT is property-type-neutral. The discount reduction applies to post-cut-off purchases regardless of what you buy — mechanics in our CGT guide — so it doesn't move the units-vs-houses needle either way.
• SMSF buyers: the LRBA ban closes new borrowing for residential property from 10 August 2026, ending the geared-SMSF route into units — cash-purchase SMSFs and personal-name buyers are unaffected.

The Case Against Units — Read This Before You Buy One

Anyone selling you units without this section is selling, not advising.

The land argument is still true. A house on 500m² in an established suburb is mostly land by value, and land is the appreciating component; a 10th-floor two-bedder is mostly building, and buildings wear out. Across most 15-year windows in the Cotality and ABS index histories, houses have out-appreciated units in the same city, often substantially. The 2026 unit outperformance is real, data-backed and durable for this phase of the cycle — but it is cyclical (affordability, yields, supply) layered on a structural disadvantage. If your horizon is 15+ years and you can service the house, the land case hasn't gone anywhere.

Strata is a cost line and a governance risk you don't control. Levies of $3,000–$8,000+ per year compound over a holding period, rise faster than CPI in facility-heavy buildings (lifts, pools, gyms), and sit entirely outside your control. The sharper risk is the special levy — a five- or six-figure bill for remediation you didn't cause, voted by a committee you're one voice on. Insurance costs across strata schemes have climbed steeply in recent years and pass straight through to levies. And the cost curve steepens with age: a building's maintenance burden compounds as lifts, membranes and services reach end-of-life, so an ageing tower with a thin sinking fund is a special levy waiting for a date.

Defects are not a solved problem. The high-rise construction boom of the 2010s produced the failures everyone remembers — flammable cladding, structural cracking, waterproofing — and the remediation tail is still running. Newer buildings are better regulated (NSW's building-commissioner regime, notably), but "new tower, unknown builder, no defect history" remains a risk profile, not a selling point.

Supply cuts both ways — eventually. The thin pipeline protecting unit owners today is a phase. When rates fall and incentives bite, approvals will restart, and the suburbs that can absorb towers will get them. The 2016–19 Brisbane and Melbourne unit glut — years of flat-to-falling unit values while houses ran — is the case study in what a supply wave does to generic high-rise stock. It is also, instructively, not what happened to boutique, land-rich, owner-occupier-grade unit stock in the same cities.

Owner-occupier depth matters at resale. Houses sell into the deepest buyer pool in the country. Investor-grade units — small, identical, in investor-dominated towers — sell mostly to other investors, a pool that evaporates whenever credit tightens or tax settings change. The stock with thin owner-occupier appeal is the stock that falls furthest and recovers last.

How to Pick a Unit That Deserves the Momentum

Our filter, in the order we'd apply it:

1. Small scheme over tower. Blocks of 4–12 have lower levies, simpler governance, fewer defect surfaces and a land share per unit that towers can't match. Villas and townhouses sit at the top of this hierarchy.
2. Land-to-asset ratio. The more of the purchase price attributable to the underlying land, the more the asset behaves like a house. A quarter-acre block cut into six villas holds serious land value per door; a 200-unit tower on the same land does not.
3. Owner-occupier majority. Ask the strata manager for the ratio. Majority owner-occupied buildings are better maintained, better governed, and sell into the deeper buyer pool.
4. Landlocked location. Established inner and middle-ring suburbs with heritage overlays, no development sites and strong amenity cannot manufacture competing supply. That is where the supply-squeeze advantage becomes permanent instead of cyclical. Our suburb selection checklist layers the standard filters (vacancy, infrastructure, demographics) on top.
5. Scarcity features within the building. Ground-floor courtyards, top-floor position, lock-up garages, extra bedrooms, character fabric — attributes the next tower can't replicate.
6. Then the strata forensics. Before unconditional: full strata report; sinking-fund balance against the maintenance plan; levy history and trajectory; minutes for defect discussion, litigation or special-levy talk; builder track record on anything under ~10 years old; cladding status confirmed in writing. Our due diligence checklist covers the toolchain.
7. Finance check last. Some lenders apply tighter LVR caps or postcode restrictions to high-density stock (and to anything under ~50m² internal); small-scheme, larger-format units generally finance like houses. This varies lender to lender more than most buyers expect — another reason the same purchase can be declined at one bank and approved at another.

City by City: Where the Unit Case Is Strongest

Sydney — the widest house-unit gap in the country (house median ~$1.48M vs units ~$860K on recent Cotality-derived figures; PropTrack's all-dwelling median $1,225,000). The affordability mechanism operates at maximum strength; the constraint is that entry-level unit yields, while better than houses, are still modest by national standards. Focus: middle-ring small schemes near the new transport spines. Context: Sydney investment outlook.

Melbourne — the counter-cyclical play. The only capital in annual decline (−0.9 to −1.1% on the June indices), never recovered its 2022 peak, and yet: unit yields now third-highest among the capitals, vacancy tighter than a year ago, rents +5.9%. When the recovery comes, the repriced-yield stock is positioned first. Context: Melbourne recovery guide.

Brisbane — the momentum market. Units were the growth leader through the boom (at one point +20% y/y against +14.6% for houses) and Cotality now ranks its entry-level units the nation's most expensive — evidence of demand depth, but also a warning that the easy repricing has happened. The house median above $1.07M keeps forcing buyers down the tier; 2032 infrastructure keeps the demand story alive (Olympics analysis). Watch approvals: Brisbane is where the next supply wave lands first.

Perth — strongest annual growth of any capital (+17–24% depending on index) with units outpacing even that, but momentum is fading fast on the daily index, and the June cross-index disagreement (PropTrack −0.5%, Cotality +0.7%) says treat single prints cautiously. Perth's yields remain among the best in the country; its supply elasticity (few landlocked suburbs) argues for the small-scheme filter applied strictly.

Adelaide, Hobart, Darwin, Canberra — Adelaide and Darwin still show solid growth with decent yields (Darwin the standout on both, from a small base and a thin market); Canberra's unit market carries a genuine oversupply history in its town-centre towers — the land-rich filter matters most there.

Selling It Later: The Exit Test Most Buyers Never Run

Every hold ends in a sale, a refinance or an estate — and all three price the asset through the eyes of the next buyer. Run the 2031 exit scenario on anything you buy in 2026:

• Who buys a house? Owner-occupiers dominate, and they buy on emotion and family logistics, in every credit environment, often above fair value. That demand floor is why house prices in established suburbs are resilient even when investor credit dries up.
• Who buys your unit? In a small owner-occupier building: downsizers, first-home buyers, young professionals — a broad pool that behaves like house demand. In a 200-unit investor tower: mostly other investors, whose appetite switches off with every credit tightening, tax change or oversupply headline. Same suburb, same year, completely different exit.
• What do the identical-stock dynamics do? In a tower, your two-bedder competes at sale with every other two-bedder in the building — including the three listed that month by other investors, and the developer's next building across the road. Differentiated stock (courtyard, top floor, character block) never faces that simultaneous-listing problem.

The current market data makes the point in real time: days on market are stretching (~28 days and rising per Cotality) and vendor discounting is widening as the downturn runs — but the widening is concentrated in exactly the segments with the thinnest buyer pools. Liquidity risk hides during booms, when everything sells in three weekends, and reprices violently in downturns. Buying stock that stays liquid in a soft market — which for units means the small-scheme, owner-occupier-grade filter yet again — is the cheapest insurance this asset class offers. Our exit strategy guide covers the sell-refinance-hold decision itself.

When a House Is Still the Unambiguous Answer

The honest scorecard by objective:

• Maximum long-run capital growth, comfortably serviceable → house, land-rich suburb. The 15-year land case stands; our cashflow vs growth framework works this trade-off properly.
• Value-add optionality → house, always. Renovation and extension upside, subdivision potential on larger or corner blocks, a granny flat for dual income, rezoning and future redevelopment plays, and knock-down-rebuild value where the land is worth more than the improvement — none of it exists under strata. A unit's value-add ceiling is a kitchen, a bathroom and flooring; a house on the right block is a pipeline of options.
• Cash flow and serviceability now, growth cycle exposure → the 2026 unit case as this article has laid it out.
• First investment on a constrained deposit → usually the unit tier by necessity — and the $184,000 entry (deposit + costs) versus $250,000 for the median house is often the difference between buying in 2026 and waiting until 2028. Deposit mechanics in our deposit guide.
• Regional markets → the calculus flips: regional house medians (~$723,000, still at record highs and +9.5% annually per PropTrack) sit below capital-city unit medians, so the affordability argument favours the regional house — land case intact, no strata. Our regional investment guide covers the trade-offs.

Which property type fits which investor

Source: our analysis, applying the framework in this article. Individual circumstances — income, tax position, existing land tax exposure — can flip any row.

Bottom Line

For the first time in years, the units-vs-houses data has a clear answer for this cycle: units are outperforming — +6.7% against +5.6% annually, smaller falls through the downturn, structurally higher yields in the first yield-expansion phase since 2023, a collapsing competing pipeline, and from mid-2027 a tax regime that channels negatively-geared money toward new attached stock. The drivers are measurable and none of them reverses quickly: borrowing capacity stays compressed while rates plateau at 4.35%, and the supply squeeze runs on a two-year construction lag.

But the long game hasn't changed hands. Land still appreciates and buildings still depreciate; a well-located house held for fifteen years remains the stronger growth asset for the investor who can comfortably carry it. The 2026 opportunity is more specific: in the price tier where most Australians actually buy, the best unit now beats the compromised house — provided you buy the right unit. Small scheme, high land share, owner-occupier building, landlocked suburb, clean strata records. That screen is the difference between owning the outperformance and owning the cautionary tale.

The practical pathway, in one pass:

• Budget under ~$850k in a capital city → shortlist boutique units and townhouses; apply the seven-step filter above before anything else.
• Objective is long-term land appreciation and you can service it → buy the house, prioritising land share and block optionality over the dwelling.
• Buying new for the post-2027 tax settings → compare the new unit against a new townhouse on land share, and model whether the developer premium exceeds the gearing-plus-depreciation benefit before committing.
• Buying purely for yield → boutique units with strong owner-occupier appeal in sub-1.5%-vacancy suburbs; check net yield after levies, not gross.