NDIS Register Integrity: 58.6% Not Currently Active
The NDIS Quality and Safeguards Commission publishes a public register of providers. In the snapshot dated 15 May 2026, that register lists 79,131 outlet rows, which resolve to 26,468 unique providers (one record per ABN). For 15,503 of those 26,468 providers (58.6 per cent), the register's own Active flag is not set to active. In other words, a clear majority of the unique providers on the national register are not currently shown as active registrations.
That figure comes from the Commission's own data, not from any judgement of ours. This article sets out exactly what it means, what it does not mean, and how it sits alongside two other measures that are easy to confuse with it.
What the register's own flag shows
Each record in the Commission register carries an Active flag. Counting one record per unique ABN in the 15 May 2026 snapshot, the split is:
So 10,965 records (41.4 per cent) carry the active flag, and 15,503 records (58.6 per cent) do not. The percentages are of the 26,468 unique providers in the snapshot. RefDat's NDIS Trust Index applies its own label, "Not Operating", to the records without the active flag, and it is those 15,503 records that this article is about. The label is ours; the underlying flag is the Commission's.
The split is reproducible from RefDat's archived copy of the register file, and the same 10,965 active count drives the live site: the provider sitemap, which lists only the providers we treat as currently active, holds exactly 10,965 URLs. That sitemap is generated from the same data, so it corroborates the count rather than confirming it from a separate source.
What "not currently active" does and does not mean
This is the part that matters most, because the headline figure is easy to over-read.
The Active flag records one thing only: whether a provider is currently shown as an active registration in the Commission's published register. A record without the active flag may be lapsed, voluntarily withdrawn, suspended, revoked, or otherwise not listed as active in this snapshot. That is the whole of what the flag tells us.
It is worth being just as clear about what the flag does not tell us:
- It is not a measure of billing. The published register does not contain per-provider payment data, so being not currently active is not evidence that a provider did, or did not, bill the scheme.
- It is not proof of fraud or misconduct. A record without the active flag is not an enforcement record. Some of these records carry a compliance action such as a suspension or revocation, and others simply show no current active registration with no specific reason in the public fields. We do not put a number on either group, because the data we use does not let us split them reliably.
- It is not proof that a provider never delivered services, or that it has closed for good. Registrations lapse and are renewed. Providers withdraw voluntarily. A not-active record is a point-in-time status, not a life sentence.
We have also corrected our own language on this point. An earlier internal definition described the active flag as meaning a provider had received an NDIS payment recently. That was wrong, and we revised it on 6 June 2026: the flag is register status only, not payment activity. We mention the correction here because the distinction is the whole basis for reading this data honestly.
A different question: do these providers have a public footprint?
Alongside the register flag, RefDat runs a separate check of its own: does a registered provider have a working website with disability-related content? This is our independent assessment, using our published methodology. It is an opinion about online presence, not a government fact, not a fraud finding, and not a judgement that a provider is unsafe. It answers a different question from the Active flag.
On that website measure, 15,563 of the 26,468 registered providers (58.8 per cent) have no meaningful public footprint: 12,343 have no website we could find, 1,818 have a website that does not load, and 1,402 have a site with no disability-services content. The detail is in our ghost providers analysis.
The two numbers, 15,503 not currently active and 15,563 with no public footprint, look almost identical. That closeness is a coincidence. They measure different things, and they are not the same providers. A provider can carry the active flag and still have no website. A provider can be not currently active and still have a perfectly good website left online. Adding the two figures, or treating one as confirmation of the other, would be a mistake, so we keep them apart.
The active providers, by transparency band
For the 10,965 providers that do carry the active flag, the Trust Index assigns a transparency band from Excellent to Poor. These bands are RefDat's assessment, not a government rating. The distribution, as a share of those 10,965 active providers, is:
| Trust Index band | Active providers | Share of active |
|---|---|---|
| Excellent | 2,661 | 24.3% |
| Good | 3,213 | 29.3% |
| Fair | 1,916 | 17.5% |
| Poor | 3,175 | 29.0% |
| Total active | 10,965 |
The band counts are exact and sum to 10,965. The shares are rounded to one decimal place; as shown they sum to 100.1 per cent, while the unrounded shares sum to 100. The denominator here is the 10,965 active providers, not all 26,468 records, because the Trust Index only bands active providers. Expressed instead as a share of all 26,468 register records, as some other pages on this site do, the same counts read Excellent 10.1 per cent, Good 12.1 per cent, Fair 7.2 per cent and Poor 12.0 per cent. We flag the denominator deliberately, because quoting a band share against the wrong base is one of the easiest ways to mislead with this dataset.
A transparency gap worth naming
The register has limits that are worth stating plainly. One example: as at May 2026, the Commission's separate compliance register lists 543 individuals who have been banned but who have no ABN attached to the banning record. With no ABN, those bans cannot be joined to a provider record by ABN in this dataset. That is not a failing of any one provider. It is a gap in how enforcement data and provider records are linked, and it is the kind of thing a public, structured register should make easy to follow rather than hard.
Why the registered population is only part of the picture
One more piece of context matters for reading any provider-register figure. Registration is not the same as operating. According to the National Disability Insurance Agency (Quarterly Report to Disability Ministers Q2 2025-26, Table E.93), 277,161 providers were paid by the scheme in the quarter to 31 December 2025, and only 17,982 of them (6.5 per cent) were registered providers. The overwhelming majority of providers paid by the NDIS are not registered at all.
The Commission register, the dataset behind this article, covers the registered side only. The two NDIA counts are internally comparable, but they come from a different body, a different date, and a different definition from the Commission register, so neither should be subtracted from our 26,468. We cite them only to make the point that the registered register is a defined slice of a much larger provider market, and that being on it, or not being active on it, is a narrower fact than it first appears.
What we are claiming, and what we are not
To be exact about the line between fact and assessment in this piece:
- Fact, from the government data: 15,503 of 26,468 unique provider records (58.6 per cent) do not carry the Active flag in the 15 May 2026 register snapshot.
- RefDat assessment, from our methodology: 15,563 registered providers (58.8 per cent) have no meaningful public footprint, and the transparency bands above for active providers.
- Not claimed: any statement that a specific provider committed fraud, billed improperly, or never delivered services. The public register does not support claims like that, and we do not make them.
A register where most records are not currently active is not, by itself, a scandal. It may simply reflect how registration lapses, renews, and turns over. But it does mean that anyone using the headline count of "registered providers" as a measure of the active market is counting something quite different from what they think. The data is public. The honest reading of it is the point.
Sources
- NDIS Quality and Safeguards Commission, provider register (Provider Finder dataset), snapshot dated 15 May 2026, retrieved 17 May 2026. Active and not-active counts reproduced from RefDat's archived copy of the published file.
- NDIS Trust Index, scoring methodology and ghost providers analysis (RefDat's independent assessment).
- National Disability Insurance Agency, Quarterly Report to Disability Ministers Q2 2025-26, as at 31 December 2025 (Supplement E, Table E.93). Provider payment counts.
Right of reply: Any provider referenced on the NDIS Trust Index is welcome to submit additional information or corrections through the right-of-reply form on their provider page. We publish verified corrections and update scores accordingly. The goal is accuracy, not accusation.