Skip to content
SDA Atlas
SDA Atlas — SDA Atlas: Australia's Specialist Disability Accommodation data
https://sdaatlas.com.au/ · Figures as at 30 June 2026
Contains data © National Disability Insurance Scheme Agency 2013, licensed under CC BY 4.0. SDA Atlas is independent of, and not endorsed by, the Agency.

SDA Atlas: Australia's Specialist Disability Accommodation data

National snapshot

As at 30 June 2026 · 7 measures

People

Active participants

782,013 ▲ 1.0% on the quarter

Dec 2019 to Jun 2026 ≥

Approved for SDA

i
25,658

65% of them are living in SDA today

Series begins Jun 2026

Living in SDA

16,644 ▲ 2.3% on the quarter

Sep 2024 to Jun 2026 ≥

Eligible, not yet using

9,014 ▼ 3.8% on the quarter

Sep 2024 to Jun 2026 ≥

Housing

Enrolled dwellings

14,235

50% of it is High Physical Support

Series begins Jun 2026

Enrolled resident places

i
31,065 est

built for 2.18 residents each, on average

Series begins Jun 2026

Pipeline (unenrolled new builds)

i
6,544

62% of it is High Physical Support

Series begins Jun 2026

Sparklines show the quarterly series where the NDIA publishes one. Its supply tables are snapshots that each release overwrites, so SDA Atlas keeps its own archive from Jun 2026 onward. Hover any figure for its source table.

Housing mix and assessed need

What is built, and what people are assessed as needing

Every participant approved for SDA is assessed for a design category, meaning the level of physical accessibility or support their home has to provide. Enrolled dwellings carry a design category too, so the two can be set against each other.

Housing is counted in dwellings and need is counted in people, so the only sound comparison is between shares: what proportion of the stock is each category, and what proportion of assessed need is for it. On that basis High Physical Support is 58% of the housing and 25% of the need. Fully Accessible housing runs the other way, and is 3.1 times rarer in the stock than it is in the need.

SHARE OF HOUSING, SHARE OF NEED HOUSING VS NEED 0% 15% 30% 45% 60% High Physical Support Housing that has been designed to incorporate a high level of physical access provision for people with significant physical impairment and requiring very high levels of support. 7,173 dwellings 5,116 people 58.3% 24.8% 2.4× more High Physical Support: 58.3% of the housing stock, 24.8% of assessed need housing than need Robust Housing that has been designed to incorporate a high level of physical access provision and be very resilient, while reducing the likelihood of reactive maintenance and reducing the risk to the participant and the community. 1,838 dwellings 1,754 people 14.9% 8.5% 1.8× more Robust: 14.9% of the housing stock, 8.5% of assessed need housing than need Improved Liveability Housing that has been designed to improve 'liveability' by incorporating a reasonable level of physical access and enhanced provision for people with sensory, intellectual or cognitive impairment. 2,066 dwellings 7,326 people 16.8% 35.4% 2.1× less Improved Liveability: 16.8% of the housing stock, 35.4% of assessed need housing than need Fully Accessible Housing that has been designed to incorporate a high level of physical access provision for people with significant physical impairment. 1,231 dwellings 6,471 people 10.0% 31.3% 3.1× less Fully Accessible: 10.0% of the housing stock, 31.3% of assessed need housing than need
Share of enrolled dwellings Share of participants with an SDA need

Shares are renormalised over the four design categories the NDIA reports on both sides. Excluded: 13.5% of dwellings (Basic and Multi-Design Category, which the demand tables do not use) and 19.4% of participants (design category not recorded). Supply from Supplement P Table P.5, need from Table P.10.

Capacity built against people housed

The NDIA publishes no occupancy figure, and no register of which dwellings are tenanted. What it does publish is enough to bound the question, and the bound is wide enough to be worth stating carefully.

Implied occupancy

53.6% est

Places without a resident

14,421 est

16,644 participants are recorded as living in SDA. The enrolled stock is built to hold about 31,065, counting every dwelling at the maximum number of residents it is enrolled for. In plain terms: for roughly every two places this housing could hold, one has a recorded resident. That is not the same as sitting empty. Nothing published says what the difference is made of: it will include new homes not yet tenanted, rooms in shared houses that suit nobody currently eligible nearby, and residents the data has not caught up with, in proportions the NDIA does not report.

How 31,065 places is calculated

The NDIA publishes dwellings grouped by the maximum number of residents each is enrolled for. Multiplying each group by that maximum gives the total number of people the enrolled stock is built to hold.

Enrolled dwellings by maximum residents, multiplied to give resident places
Enrolled for Dwellings Places
1 resident 5,705 5,705 × 1 = 5,705
2 residents 4,112 4,112 × 2 = 8,224
3 residents 2,169 2,169 × 3 = 6,507
4 residents 767 767 × 4 = 3,068
5 residents 1,331 1,331 × 5 = 6,655
6 or more residents * 151 151 × 6 = 906
Total 14,235 31,065

* The top band is open ended, so counting it as six makes the total a floor rather than an exact figure. Source: Supplement P Table P.6. Capacity is what a dwelling is enrolled to hold, not how many people live in it.

Read this one carefully. Implied occupancy is an SDA Atlas estimate, not an NDIA figure. It divides participants recorded as living in SDA by enrolled resident places, which come from a different NDIA table. The NDIA notes that enrolled capacity “may not accurately reflect current SDA participant occupancy”. It is not a vacancy rate.

Lowest implied occupancy

SA4 regions with 100 or more enrolled dwellings

  • Mandurah WA 23%
  • Perth South West WA 33%
  • South Australia South East SA 33%
  • Townsville QLD 35%
  • Perth North East WA 36%
  • Perth South East WA 36%

Low implied occupancy is not the same as oversupply. Places can stand empty because the homes built in a region are not the kind that the people there were assessed as needing. The NDIA names a different set in the SDA Report 2026 (p.6) as showing higher supply relative to identified need: Melbourne - West, Adelaide - North, Ipswich, Adelaide - South, Logan - Beaudesert. None of them appears above.

The scheme over time

Active participants since Dec 2019, from the longest series the NDIA publishes at regional granularity.

0 250k 500k 750k 1m Dec 2019 Jun 2021 Dec 2022 Jun 2024 Jun 2026 781,982
Active participants

Living in SDA, and eligible for it

0 5k 10k 15k 20k Sep 2024 Jun 2025 Jun 2026 16,596 8,985
Living in SDA Eligible, not yet using

Summed from district figures, so these end slightly below the tiles above: the NDIA suppresses district counts under 11, and those cells cannot be added in. The gap is the suppressed districts, nothing else.

Supported Independent Living

0 10k 20k 30k 40k Mar 2019 Dec 2021 Jun 2024 Jun 2026 36,864
SIL participants

SIL funds the support a person receives at home; SDA funds the home itself. Many people are funded for both.

Go deeper

The SDA gap

Where the 9,014 people found eligible for SDA and not living in it actually are, district by district.

The housing stock

What has been built, and of what kind, by design category, building type and resident capacity.

Where it is

Supply, need and implied occupancy across every statistical region in the country.

What is coming

6,544 dwellings are registered and not yet enrolled, and why that is not the same as under construction.