Best horse racing tipsters UK 2026

Every active PuntHub tipster, ranked by verified return on investment. Every figure on this page is calculated from the public bet ledger — the same ledger you can read, bet by bet, on each tipster's own page.

Last updated 6 September 2026 · 3,133 bets on record

The table

Ranked by lifetime ROI. Profit is to level stakes in points, so it is comparable across services regardless of what you stake per bet. Sample size and date range are shown for every row — a 30% ROI over 40 bets is not the same claim as 12% over 4,000.

#TipsterSport / typeBetsStrike rateROIProfit (level stakes)Record covers
1AI Lays
AI-powered lay betting on Betfair Exchange — verified results, full automation.
horses1,83379.8%10.0%+183.1 ptsOct 2024 – Sept 2026
2Draw Specialist
Football draw predictions — 5 years of data, 10-15% ROI.
football20531.7%6.2%+12.6 ptsAug 2025 – Dec 2025
3Value Engine
The AI that hunts mispriced winners — free public trial running.
horses39527.1%3.0%+12.0 ptsAug 2025 – Sept 2026
4Andys Lays
Expert lay selections from one of the industry's best.
horses70078.7%2.5%+17.7 ptsSept 2025 – Sept 2026

Looking for the full directory rather than the ranking? Browse all tipsters — or see today's free horse racing tips.

How does PuntHub verify a tipster?

Every selection is written to the bet ledger before the race starts, with the date, time, event, selection, stake and the odds available at the time of posting. When the race settles, the result and profit or loss are written against that same row. Rows are never edited after settlement and never deleted, so a losing month cannot be quietly removed. The figures in the table above are recalculated from that ledger every time this page is regenerated — they are not typed in by hand and there is no separate marketing number.

What does ROI actually mean here?

ROI is profit divided by total points staked, expressed as a percentage. A tipster with a 12% ROI has returned 12 points of profit for every 100 points staked. It is the fairest single measure for comparing services that bet at different odds and different volumes, which is why the table is ranked on it rather than on raw profit — a service that has been running for five years will always show more total profit than one that started last spring, regardless of which is better.

Why does sample size matter?

Because a small sample tells you almost nothing. Forty bets is a run of luck; four thousand is a method. Every row shows the bet count and the date range the record covers so you can weight the ROI accordingly. Be sceptical of any tipster — here or anywhere else — who publishes a headline percentage without telling you how many bets it is built on and over what period.

Is a high strike rate a good thing?

Not on its own. Lay betting services win a high proportion of their bets by design, because they are betting against short-priced runners — but a single loss can cost several wins. Backing services at bigger odds win less often and lose money for long stretches before a winner lands. Read strike rate and ROI together, never separately.

Does ranking first mean it is the right one for you?

No. The table ranks on ROI alone. It cannot know your bankroll, whether you can place bets during the working day, or whether you are comfortable with the liability that lay betting carries. A service posting one selection a day suits a very different punter from one posting fifteen. Read the individual tipster pages before you subscribe — each one carries the full bet history, the pricing and the cadence.

Past performance is not a guarantee of future results. Nothing on this page is betting advice. 18+ · BeGambleAware.org