How we pick radiators

Where the numbers come from

Every plumbed radiator on this site carries its manufacturer-published heat output at Δt50 — the 75/65/20°C test condition that BS EN 442 defines and that the UK trade quotes as standard. We take those figures from the maker’s own data sheet and link it on the product card so you can check us. We do not re-test radiators and we do not estimate outputs: if a manufacturer hasn’t published a figure, the radiator isn’t listed.

For electric and oil-filled radiators the output is simply the wattage — an electric heater turns essentially all the electricity it draws into heat in the room, so 1 W is 3.412 BTU/hr. That makes the interesting questions about control (thermostat accuracy, timers, occupancy sensing) rather than raw output, which is what our write-ups focus on.

How the ranking works

Each radiator gets a score out of 1 from six weighted parts:

Shortlists then show the size a typical room of that kind needs, at the height people actually buy, rather than whichever size happens to score best.

Where commission fits — and where it doesn’t

Commission is not one of the six factors. It is used in exactly one place: when the same radiator is sold by several retailers, we link to the one we have an affiliate relationship with. That changes which shop you land on, never which radiator we recommend. Links that earn us something are marked as sponsored in the page code.

Right now 46 of our 333 listings carry a commission-earning link and the rest earn us nothing at all — we list them because they’re the right answer, which is rather the point.

What we can’t tell you

• We have not physically tested these radiators. Our judgements come from published specifications, manufacturer data and how the product is built — not a lab.
• Prices move. Where we show one it is the retailer’s price when we last refreshed the feed; always check before buying.
• Heat-pump outputs are a real weak spot across the whole industry. Most makers publish Δt50 prominently and the low-temperature figures reluctantly. Where we flag heat-pump suitability we are reasoning from construction, not from a published Δt30 number, unless we say otherwise.
• Reader click-through tells us what looks appealing, not what performs well once fitted. It is weighted at 20% for that reason.

Telling us we’re wrong

If a figure here disagrees with a manufacturer’s current data sheet, the manufacturer is right and we want to know. The data sheet we used is linked on every card.