£26.6bn, £31.1bn, £37bn, £65bn — four figures in circulation for the same thing, quoted as fact by different outlets. The system meant to settle the question hits a legal deadline in eleven weeks. Savers still will not be able to use it.
There is a large amount of money sitting in UK pension pots whose owners have lost track of them. How large is genuinely unclear, and the range of published estimates is wider than most of the individual figures.
Search for the total and you will be told it is £31.1bn. Or £65bn. Or £26.6bn. Or £37bn. Each appears in reputable coverage, stated without qualification. The gap between the lowest and highest is £38.4bn — larger than the lowest estimate itself.
The averages diverge just as sharply. Depending on the source, the typical lost pot is worth £9,470, or £19,500, or around £23,000.
These are not competing measurements of one thing. They are measurements of different things, and almost nobody says so.
Where the numbers come from
£31.1bn across 3.3 million pots is the Pensions Policy Institute figure, produced for the Pension Attention campaign and the most widely cited. It gives an average of £9,470, rising to £13,620 among savers aged 55 to 75. The PPI reported this was up around 60%, or nearly £12bn, since 2018.
£26.6bn across 2.8 million pots is the PPI’s earlier estimate from 2022. It still circulates on financial websites as current. The arithmetic gives an average of about £9,500 — consistent with the later figure, which suggests the population grew rather than the methodology changing.
£37bn across 1.6 million pots came from research by Profile Pensions. That works out at roughly £23,000 a pot. Half as many pots as the PPI counts, but a higher total.
£65bn, average around £19,500, is attributed to MoneySavingExpert and has circulated widely.
The pattern is visible once the figures are set against each other. The estimates producing low averages count many small pots; those producing high averages count fewer, larger ones. That points to different definitions of what counts as lost.
A pot can be genuinely lost — the provider cannot contact the owner and the owner does not know it exists. It can be dormant — no contributions, but the owner knows about it. It can be inactive but perfectly well documented. Sweep in all three and you get a big number. Count only the first and you get a smaller one.
None of the published figures is a count. They are all estimates, extrapolated from surveys and provider samples. There is no register of lost pensions to check against, which is precisely the problem being measured.
Why the register does not exist yet
The intended fix is the pensions dashboard: a service showing all of a person’s pots in one place, drawn from a central data system that providers connect to.
The legal connection deadline is 31 October 2026 — eleven weeks from now. It is set in the Pensions Dashboards (Amendment) Regulations 2023, which define it as the latest date by which relevant occupational schemes must be connected to the Money and Pensions Service. It applies to schemes with more than 100 members.
Progress is substantial. The Pensions Dashboards Programme reported in April that more than 1,000 providers and schemes and 60 million private pension records had connected. By June, coverage was reported at around 85%, with the final month-end deadlines applying to schemes of 100 to 249 members. Smaller schemes are not obliged to connect but may opt in.
Trustees who miss the deadline face penalties reported at up to £5,000 for an individual trustee and £50,000 for a corporate trustee.
The gap nobody mentions
Here is the part that gets lost in the coverage: the connection deadline is not the launch date.
31 October 2026 is when the pipes must be connected. The consumer-facing MoneyHelper dashboard — the thing an ordinary saver would actually use to find a forgotten pot — is not expected before the 2027/28 financial year.
So the deadline widely reported as the moment lost pensions get solved is, for savers, not a moment at all. Nothing becomes visible to the public on 1 November. The wait continues for at least another year after that.
There is also a data-quality question sitting underneath. In May, the Pensions Regulator launched a targeted assessment of around 240 defined benefit and hybrid schemes whose value data was judged more likely to be out of date than other scheme types. Connecting a scheme to the system does not by itself guarantee the figures it supplies are current.
Meanwhile, hardly anyone looks
Research by the tracing service Gretel found around 44% of UK adults suspect they hold a lost or dormant account, while only 13% have ever tried to trace one. That is roughly one person in three who thinks money may be theirs and has never checked.
Use of the free government service is rising. Department for Work and Pensions figures indicate about 834,000 people used the Pension Tracing Service during 2025, and analysis by the accountancy firm Lubbock Fine found calls rose by more than half in the year to September.
Set against 3.3 million estimated lost pots, that is still a small fraction.
What you can do now, for free
None of this requires waiting for the dashboard, and none of it costs anything.
The government’s Pension Tracing Service is free and does not require you to remember your provider — an employer name and approximate dates are enough. It returns contact details for the scheme, and you then contact them yourself.
Before that, go through old paperwork. Payslips show pension deductions and often name the scheme. So do P60s and old contracts. An employer name and rough dates are usually sufficient.
Do not dismiss short jobs. Auto-enrolment means even a few months of work can have created a pot, and those small pots are exactly the ones people forget. Multiple modest pots are the common pattern, not one large one.
Be wary of paid tracing services. The government service is free, and firms charging a percentage of what they find are recovering your own money for a fee.
If you find a pot, tracking it down is not the same as deciding what to do with it. Whether to consolidate several pots into one depends on charges, and on whether any older scheme carries guarantees — such as a guaranteed annuity rate — that would be lost on transfer. That is a question for regulated advice, or at minimum for MoneyHelper’s free guidance, not for a news article.
The figure to watch
When the dashboard eventually reaches savers, it will for the first time make the real number knowable. Not an extrapolation from a survey — an actual count of pots that connect to a person and pots that do not.
If that number lands far from £31.1bn in either direction, it will say something about how the estimates have been constructed for the past decade. That is the moment worth waiting for, and it is further away than the October deadline suggests.
