The State of Disposable Email, September 2026

As of September 6, 2026, we track 217,847 disposable email domains, and a live Domain Name System (DNS) sweep run the same day found that 35.91 percent of them can still receive mail. New throwaway domains kept arriving on every single day of our 36 day observation window, 10,223 in all, at a steady rate almost identical to last month's. And for the first time, this report can measure domains dying, and coming back to life, because the recheck history we promised in August is now collecting.

Headline findings

• We track 217,847 disposable email domains, and only 35.91 percent of them can still receive mail today.
• 10,223 genuinely new disposable domains appeared in our 36 day window, and the steady discovery rate was 185.1 per day, next to 185.7 in August. The flow is not slowing down.
• Of those 10,223 new domains, the widely bundled mailchecker blocklist contained exactly 1.
• The most complete free list we tested misses 63.07 percent of the disposable domains that are live right now.
• Inside the window, 1,686 tracked domains went from alive to dead and 1,215 came back the other way, the first churn we have measured in both directions.
• .xyz supplied 11.40 percent of the window's new domains, more than double its 4.72 percent share of the full dataset.

How many disposable email domains exist right now?

We currently track 217,847 disposable email domains, aggregated from six upstream sources and refreshed daily. That total is the accumulated picture, not a count of domains that work today. The same day sweep found 78,224 domains, 35.91 percent, that could still receive mail.

This is the second monthly edition of this report, so the numbers now have somewhere to move from. Set against the August 2026 edition:

Metric	August 2026	September 2026
Domains tracked	211,812	217,847
Alive today	73,110 (34.52%)	78,224 (35.91%)
New in the observation window	10,864 in 34 days	10,223 in 36 days
Steady daily average, intake events excluded	185.7	185.1
Most complete free list, live domains missed	61.16%	63.07%
mailchecker, window's new domains contained	4 of 10,864	1 of 10,223

The dataset grew by just over six thousand domains in a month, the live share ticked up as fresh, working domains arrived faster than old ones died, and the underlying discovery rate barely moved. One month is not a trend, but the stability of that steady rate is the number to watch.

How fast do new disposable domains appear?

Constantly, and at almost exactly last month's pace. In the 36 days from August 1 to September 5, we detected 10,223 genuinely new disposable domains, an average of 284.0 per day with a median of 177.5. One intake event sits inside that total: a three day surge in one of our upstream lists delivered 4,115 domains between August 13 and 15, peaking at 2,759 on August 14. Set those three days aside and the remaining 33 days averaged 185.1 genuinely new domains per day, effectively unchanged from August's 185.7. No day in the window produced zero.

New disposable domains detected per day across the observation window. The one time bulk import that seeded the dataset is excluded.

These new arrivals are overwhelmingly real, working domains: 9,971 of the 10,223 still pass MX validation today, with 237 already dead and 15 unknown. Grouping them by the mail infrastructure they route through shows the same clustering as last month with a reshuffled head: mail2world.com received 1,253 of the window's new domains and catchmail.io 1,212, with cleantempmail.com at 546 and cloudflare.net at 341 behind them. The reading caution from August still applies: a cluster names shared mail infrastructure, not necessarily a disposable operator, and several of these are commodity mail hosts that operators simply route through. 460 of the window's new domains have no mail host today and sit outside any cluster.

Which top level domains do disposable providers use, and why does the concentration matter?

The historical mass still concentrates in a few top level domains (TLDs). Across all 217,847 domains, the top five TLDs hold 121,229, 55.65 percent of the dataset, out of 680 distinct TLDs, and .com alone accounts for 81,358 domains, 37.35 percent. One change to the leaderboard this month: .xyz overtook .net for third place, 10,286 domains to 9,878.

Share of all 217,847 tracked disposable domains, by top level domain.

The concentration remains economic: the roster pairs abundant, unremarkable .com with the discount new gTLDs and the historically free country codes, and cheap registration is what makes rotating through domains nearly free. The month's real story is in the new detections, where the mix shifted hard.

Top level domain mix of newly detected disposable domains versus the full dataset.

.com ran hot at 46.16 percent of new detections against 37.35 percent of the corpus, and .xyz ran at 11.40 percent of new against 4.72 percent overall, more than double its historical weight. .uk stayed elevated at 2.73 percent of new versus 0.58 overall, .pro appeared at nearly four times its weight, and .ru continued to fade, 2.49 percent of new against 5.32 of the corpus. In August the overweight TLDs were .uk and .id; in September they are .xyz and .pro. Where new disposable domains concentrate moves month to month, which is exactly why a detector tuned to the historical distribution lags the present.

How much of the disposable ecosystem is already dead?

Most of it, though slightly less than last month. The report day sweep classified every tracked domain by whether mail could reach it.

Status	Domains	Share of all 217,847
Alive, mail could be delivered today	78,224	35.91%
Dead, no mail route today	136,067	62.46%
Unknown, transient DNS failure	3,556	1.63%

Of the live domains, 51,785 publish working MX records and 26,439 qualify only through the RFC 5321 address record fallback. Of the dead, 132,703 have no mail route at all and 3,364, 1.54 percent of the dataset, publish an explicit null MX. The unknowns hit transient DNS failures through two retries and count as neither alive nor dead.

The two standing cautions travel with these numbers. The dead share describes the accumulated corpus, mostly historical entries that have lapsed, not a death rate. And dead entries are never harmless to carry, because an expired domain re-registered by a legitimate owner turns every stale blocklist entry into a false block. This month, for the first time, we can put early numbers on both directions of that churn.

How fast do disposable domains die, and come back?

We can begin to answer this honestly now, with limits stated up front. In August we said per domain lifespan could not be computed because no recheck history existed; that history began collecting on August 9 and now covers 215,754 domains, each re-tested on a rolling schedule with every status change recorded.

The strict 30 day mortality figure is still CANNOT COMPUTE for this window: no domain has been under observation for a full 30 days yet, and the earliest first observation makes the figure computable for windows ending on or after September 9, which means the next edition. What the young history already shows is observed decay: of the 77,893 domains first seen alive and rechecked at least once since, 1,431, which is 1.84 percent, were dead or null MX at their latest recheck, over a median observation span of just 17 days. Read that as an early floor, not a lifespan estimate.

The window's status changes are the more striking early result, because they run both ways: 1,686 tracked domains went from alive to dead between August 1 and September 5, and 1,215 came back from dead to alive. Domains do not just die, they resurrect, as parked names get repointed and lapsed infrastructure returns. Every one of those 2,901 transitions is a domain about which any static snapshot became wrong within five weeks, in one direction or the other.

How well do the popular free blocklists actually perform?

We downloaded fresh copies of the same three community blocklists on report day and judged them against the same live MX sweep. All three remain real, maintained projects; two of them shipped commits within days of our measurement, and the gap is structural rather than a maintenance failure.

Blocklist	Entries	Live disposables it misses	Dead weight	New in our window it caught
disposable-email-domains	8,725	71,721 (91.69%)	2,119 (24.29%)	518 of 10,223
disposable/disposable	75,260	49,333 (63.07%)	45,594 (60.58%)	763 of 10,223
mailchecker	56,356	59,977 (76.67%)	37,492 (66.53%)	1 of 10,223

Free blocklist coverage gaps and dead weight, both measured against a live MX sweep of every domain run the same day.

The shape matches August, with the edges slightly sharper. Coverage: the most complete list, at 75,260 entries, misses 49,333 of the 78,224 live domains, and the most widely copied one misses 91.69 percent. Freshness: of the window's 10,223 new domains, the best list contained 763 and mailchecker contained 1, down from 4 last month. Dead weight still runs as high as 66.53 percent, and this month's resurrection data gives that risk a second face: 1,215 domains came back to life inside the window, and a list that recorded any of them as dead is now blocking or excusing the wrong thing. In fairness, the reverse comparison also still holds: 85 domains appear in at least one of the three lists but not in ours, every one an exclusion our pipeline makes deliberately because it honors allow lists that protect relay services and other legitimate domains.

What does a null MX record mean?

A null MX record is a domain declaring, in DNS, that it never accepts email: a single MX record pointing at the root, defined in RFC 7505, and the cleanest no-mail signal that exists. In this month's dataset, 3,364 domains, 1.54 percent of the total, publish one. We count them among the dead because, by their own configuration, that is what they are. The full mechanics, including why a null MX is the opposite of simply having no MX record, are covered in our null MX explainer.

Why are relay services like Apple Hide My Email not disposable?

Because a relay address forwards to a real person's permanent inbox, and a disposable address does not. A relay or forwarding service, Apple Hide My Email, Firefox Relay, SimpleLogin, addy.io, DuckDuckGo Email Protection, or Proton, gives a privacy conscious, often paying user a masked address that keeps working. We track 12 relay domains across those 6 provider families, a count that reflects the providers shipped in the current product configuration rather than a live database query, and we flag them as a separate signal that is never part of the disposable verdict and never blocked by default. Twelve stable relay domains against 10,223 new throwaways in five weeks is the entire distinction in one comparison: relays are not hiding from anyone. Blocking them to stop throwaways turns away real customers.

What should you do if you run a signup form?

Name the risk correctly first. Letting throwaway signups through does not damage your sending reputation at Gmail, Outlook, or Yahoo; mailbox providers judge you on how their own users treat your mail, and a dead disposable domain has no users at any of them. The damage lands at your Email Service Provider (ESP). Accepted throwaways go dark on a delay, this month's decay data shows 1.84 percent of an alive cohort dead again within a median 17 days of observation, and when your next campaign hits them, the hard bounces arrive together. A bounce spike is one of the strongest list quality signals platforms like Mailchimp and SendGrid police, and the consequences run from warnings to forced list cleanings to suspension. That is the real cost, and it is entirely avoidable at the form.

The fix is the same as last month because the data is the same shape: validate at signup against something that moves at the flow's speed, block confirmed throwaways with a clear message, let relay addresses through, and fail open so the check never breaks the form. This report is drawn from the same live, MX validated dataset that isitdisposable.com uses to do exactly that in one API call or a drop in snippet, refreshed daily, with a free plan of 250 lookups a month and a 14-day full access trial, no credit card, if you want to see your own signup traffic through this lens.

Methodology

Every figure in this report comes from a read-only extraction of our production database on 2026-09-06 at 00:00:35 UTC, paired with a live DNS sweep of all 217,847 tracked domains run the same day. Nothing is estimated or extrapolated, and any metric we could not compute honestly is labeled CANNOT COMPUTE with the reason. The dataset is aggregated from six upstream sources and refreshed daily. Aliveness uses the same resolution rules we apply in production: an MX lookup first; an A and AAAA address record fallback per RFC 5321 when a domain has no MX record and no null MX; and the null MX convention per RFC 7505. Transient DNS failures were retried twice, and anything still unresolved is reported as unknown, never counted as dead.

The observation window is the 36 days from 2026-08-01 to 2026-09-05. The initial bulk import that seeded the dataset, 200,864 domains sharing a single load timestamp, is excluded from every growth figure and included in every snapshot figure; the extraction day is excluded from the window and had contributed no new domains by extraction time. This is the second monthly edition. The recheck job that makes lifespan analysis possible began collecting on 2026-08-09 and covers 215,754 domains; the strict 30 day mortality figure becomes computable for windows ending on or after 2026-09-09 and is expected in the next edition, alongside deeper churn analysis as the history grows. Customer and API usage statistics are not part of the public data series.

Cite this data

You are welcome to cite any statistic in this report. Please attribute it to the isitdisposable.com State of Disposable Email report, September 2026, and link to this page. If you need a figure that is not here, or a breakdown we have not published, ask, and we will see whether the data supports it.