Premium Bonds have long been one of Britain’s most beloved ways to save: no risk to your capital, a monthly “thrill” of a prize draw, and the chance—however remote—of a £1 million jackpot. But after a series of prize-rate reductions and with top easy-access savings accounts still paying around 4.5% in late 2025, the obvious question for UK savers is whether Premium Bonds are still a smart home for cash. MoneySavingExpert.com+2MoneyWeek+2
That question has also been at the centre of a recent debate in the UK press, including The Economist’s look at how “rational” the country’s soft spot for Premium Bonds really is—hinting that your tax position is a big part of the answer. The Economist
Below is a clear, news-style guide to what’s changed, what the numbers actually mean, and when personal finance experts—including Martin Lewis—say Premium Bonds can be “very good” compared with standard savings.
The big Premium Bonds reality check: “3.60%” isn’t a guaranteed interest rate
If you see headlines quoting a “rate” for Premium Bonds, it’s crucial to know what that figure is—and what it isn’t.
NS&I’s Premium Bonds currently list an annual prize fund rate of 3.60% (variable) and odds of 22,000 to 1 per £1 Bond number per month. But Premium Bonds do not pay interest in the usual sense; instead, that prize fund is distributed through monthly draws.
In other words:
- The prize fund rate is a portfolio-wide average of prizes paid out across all eligible Bonds.
- Your personal outcome can be higher, lower, or even zero, depending on luck—especially with smaller holdings.
MoneySavingExpert puts it bluntly: there’s “no guaranteed return,” and many savers can do better with normal accounts—particularly if they don’t pay tax on savings interest. MoneySavingExpert.com+1
What the December 2025 prize draw tells you about “typical” winnings
Premium Bonds are often discussed like a savings product, but their prize structure behaves more like a “safe lottery” (your stake is protected, but returns are skewed).
NS&I’s own breakdown for the December 2025 draw shows just how top-heavy the prizes are:
- 2 prizes of £1 million
- 77 prizes of £100,000
- 153 prizes of £50,000
- …but millions of smaller prizes (£25/£50/£100)
That skew matters because it means a small number of people pull the average up—while many get a modest return or none at all in any given year.
Martin Lewis’s core warning: for many savers, “best-buy” savings beat Premium Bonds
Martin Lewis (via MoneySavingExpert) has repeatedly emphasised a point that often gets lost in viral social posts and headline summaries:
If you’re a saver within your tax-free allowances, the best easy-access savings accounts and cash ISAs can often beat Premium Bonds—without the uncertainty.
MoneySavingExpert notes that while the prize rate is 3.6%, “if you have average luck you won’t win as much as the 3.6% prize rate,” and then compares Premium Bonds to a 4.5% savings account using probability modelling. MoneySavingExpert.com+1
One striking finding from their analysis (assuming you don’t pay tax on interest) is that Premium Bonds have only an 11% chance of beating a 4.5% savings account over a year even with £50,000 in Bonds.
That’s not an argument that Premium Bonds are “bad.” It’s an argument that for many people, certainty wins.
So when are Premium Bonds “very good” compared with savings accounts?
This is where the nuance—and the “it depends”—becomes real.
1) When tax makes your best savings rate much lower
Premium Bonds prizes are free from UK Income Tax and Capital Gains Tax.
Savings interest, meanwhile, can be taxed once you exceed the Personal Savings Allowance (PSA):
- £1,000 for basic-rate taxpayers
- £500 for higher-rate taxpayers
- £0 for additional-rate taxpayers
MoneySavingExpert’s comparison shows how quickly tax can erode a headline savings rate. Using its example of a top easy-access account paying 4.5%, it estimates the effective rate after tax can drop significantly—while Premium Bonds remain tax-free.
Translation: Premium Bonds can become comparatively more attractive if you’re a higher- or additional-rate taxpayer who has already used up tax shelters like ISAs and your PSA.
2) When you’ve already maxed your ISA allowance—but still want tax-free “cash-like” options
Cash ISAs remain a straightforward way to earn tax-free interest up to your annual ISA limit. But once that allowance is used, people often hunt for other tax-efficient places for cash.
MoneyWeek, citing AJ Bell and other experts, notes Premium Bonds can appeal to savers seeking tax-free options beyond the ISA, with the important caveat that you must accept the possibility of winning nothing.
3) When you value “safe upside” more than predictable income
Premium Bonds are designed for people who want:
- easy access to their money,
- no guaranteed returns, and
- the chance of tax-free prizes from £25 to £1 million.
That preference isn’t necessarily irrational—it’s just a different objective than “maximise guaranteed interest.” In behavioural economics, products like Premium Bonds are often grouped into prize-linked savings: standard savings mechanics with a lottery-like feature.
A major review of prize-linked savings by Melissa Kearney, Peter Tufano and co-authors explains why these products can be appealing: they offer skewed returns (a small chance of a big win) and an element of “fun,” while still returning the principal. Pension Research Council
4) When you have a large balance and want government backing beyond bank protection limits
This is a quieter reason Premium Bonds (and NS&I more broadly) remain popular among some savers: security.
- Bank and building society deposits are protected by the FSCS up to £120,000 per person, per authorised firm as of 1 December 2025.
- NS&I says it is backed by HM Treasury and guarantees 100% of everything you invest, “not just £120,000—every penny.” NS&I+1
For people holding substantial cash (for example, temporarily parking money), that government guarantee can be psychologically—and practically—reassuring.
When Premium Bonds are usually a poor choice
Even strong Premium Bonds fans should be honest about where they underperform.
Small pots: the “most people never win” problem
A headline statistic from MoneyWeek—based on an FOI request by AJ Bell—captures the distribution issue: around 63% of Premium Bond holders have never won a prize, and many accounts have very small balances, which naturally makes winning less likely.
MoneyWeek also quotes AJ Bell’s Laura Suter highlighting that a majority of holders have tiny balances, while a small minority hold very large sums—an imbalance that helps explain why Premium Bonds can feel “amazing” for some and “pointless” for others. MoneyWeek
If you need predictable interest for bills or budgeting
NS&I itself flags Premium Bonds are “not for you” if you want:
- “regular income”
- “guaranteed returns” NS&I
If you’re building an emergency fund and want certainty, the peace of mind of a guaranteed AER can beat the thrill of a draw.
If you’re chasing “hacks” or myths
Premium Bonds are a random draw. There’s no magic purchase pattern that boosts your odds beyond holding more eligible £1 Bond numbers.
Martin Lewis has publicly dismissed one common myth—whether buying bonds in separate chunks versus one block changes your chances—as “absolute nonsense,” saying there’s “no clever way” to influence the outcome. The Independent
MoneySavingExpert similarly notes NS&I’s ERNIE is an audited random number generator—so “systems” and “strategies” are mostly noise. MoneySavingExpert.com
Why Brits love Premium Bonds anyway: the “savings-gambling hybrid” effect
If Premium Bonds aren’t the highest-paying option for many people, why are they still so popular?
A classic analysis by Peter Tufano describes Premium Bonds as a “savings/gambling hybrid”: principal is protected, but returns are distributed like a lottery. His paper notes that sales respond both to “investment-like” variables (relative yield) and “lottery-like” variables (large prizes). ResearchGate+1
Another paper on Premium Bonds’ success argues that factors like prize skewness and the maximum holding amount play an outsized role in demand—again pointing to the emotional punch of “could win big,” even when the biggest prize contributes relatively little to overall expected value. SSRN
This helps explain why Premium Bonds remain culturally sticky in the UK: they’re not just a financial product, they’re a ritual—monthly anticipation, family gifting, and that uniquely British mix of caution (don’t lose the principal) and optimism (maybe we’ll be lucky).
A practical checklist: should you keep your money in Premium Bonds right now?
If you’re trying to decide in late 2025 heading into 2026, here are the decision points that matter most.
Premium Bonds may fit if:
- You’re a higher- or additional-rate taxpayer, and your savings interest is (or will be) taxed beyond your PSA.
- You’ve already used your ISA allowance and still want tax-free potential returns.
- You value capital security and like the option of HM Treasury backing beyond deposit insurance caps.
- You can tolerate unpredictable outcomes and are comfortable that you might win nothing.
You may be better off in a top savings account or cash ISA if:
- You’re within your PSA and can access competitive rates (around 4.5% in late 2025) with no uncertainty.
- Your Premium Bonds holding is small and you’re relying on it to “do something” each month. MoneyWeek
- You need a predictable monthly interest stream (Premium Bonds don’t offer one).
Bottom line: Premium Bonds can be rational—just not for the reasons most people think
Premium Bonds aren’t automatically a mistake, and they aren’t automatically a bargain. They’re best understood as tax-free, government-backed cash with a lottery-like return profile.
For many UK savers—especially those who don’t pay tax on savings interest—Martin Lewis’s message is consistent: the maths often favours best-buy savings.
But for higher earners facing tax on cash returns, or for people who prioritise safety and enjoy the upside optionality, Premium Bonds can still be a reasonable and even strategic choice—so long as you go in with open eyes about what the “rate” does (and doesn’t) promise. NS&I+1
