What the UK actually buys from China, and where a smaller buyer is locked out

Where these numbers come from

Everything below was queried from HMRC's own trade statistics API on 22 September 2026: UK imports from China, for the calendar year 2025, aggregated by commodity. It is the same dataset behind the official trade figures rather than an estimate, and anyone can re-run it.

One caveat stated up front. The data gives value and net mass, so the density figures below are pounds per kilogram. Freight is charged on volume as much as weight, so weight is a proxy rather than the exact measure, and a category of light bulky goods will look denser here than it feels on a quotation. It is good enough to rank categories against each other, which is what it is used for.

The shape of the trade

Total UK goods imports from China in 2025 came to £62.6 billion. The concentration at the top is extreme and not addressable by a smaller business: smartphones alone were £4.56 billion on one commodity code, portable computers £3.79 billion, and electric and hybrid cars several billion more. Those are brand-owned supply chains, not categories anybody sources into.

Below that layer the picture changes, and it is the layer worth reading. The table gives the larger chapters by value, sorted by density, because density is what decides whether scale is a barrier.

A correction to that table, and the question it answers

A reader asked the obvious next question: if bulk is the barrier, are the dense categories the better ones to ship? Checking it properly exposed a flaw in the table above, so here is the correction rather than a quiet edit.

A chapter average is dragged upwards by a few extraordinarily dense lines. Smartphones run at roughly £1,400 a kilo and are a third of the electrical chapter on their own; video game consoles run at about £88 a kilo and are a quarter of the toy chapter. Strip out the brand lines nobody can source into and the picture changes: toys fall from £6.39 to £4.93 a kilo once consoles are removed, and electrical machinery falls from £15.40 to £9.74 once smartphones and network equipment are. Optical and medical barely moves, from £19.77 to £18.87, because no single line dominates it — the largest is six per cent of the chapter.

So the honest reading of toys is the opposite of what the chapter average suggests. At £4.93 a kilo, real toys sit in the same bulky band as furniture, which makes them a scale-barrier category rather than an easy one — and they carry the heaviest consumer safety regime of anything in this table.

Use the table to rank categories, then check the composition of the one you care about before trusting the number. The query behind all of this is public and takes a minute to re-run.

Density buys you cheaper freight, not a better business

The deeper answer to that reader's question is that cheap to ship and worth importing are close to opposites, and for a reason worth sitting with.

Where freight is a small fraction of the goods value, the scale barrier is low. That is genuinely easier — a part-container shipment does not ruin the arithmetic, and an SME can enter. But so can everybody else, and a category anybody can enter has already had its margin competed down. The saving you were hoping to find is the saving a hundred other importers already found.

Where freight is large relative to value, the barrier is real, which is precisely why a gap survives there for somebody able to get to full-container economics. That is uncomfortable but it is how the numbers behave.

The barrier does not disappear in dense categories; it moves. In electrical goods it becomes brand and electrical safety. In optical and medical it becomes the regime itself: those goods are trivial to ship and among the hardest things in this table to place on the UK market lawfully, because a medical device needs a UK Responsible Person and registration before it is sold, which our medical devices and PPE guide sets out. Shipping was never what stopped anyone there.

The practical version: pick your barrier. A freight barrier can be solved with volume, patience or a shared container. A regulatory barrier is solved with work that does not get cheaper as you grow, and a brand barrier usually is not solved at all.

Why the bulky categories are the ones with a locked door

Sea freight is priced per container, so the cost per unit falls as the container fills and rises sharply when it does not. On a category at £15 a kilo the freight is a small fraction of the goods value and a part-container shipment is survivable. On a category at £2 a kilo it is not: the same freight bill is spread over goods worth a fraction as much, and a less-than-container load can wipe out the entire saving that made the import attractive.

That is the real barrier, and it is not the tariff. A national retailer buying forty containers a year of furniture is paying the efficient freight rate per unit and capturing the gap. An independent shop that needs six sofas is paying a UK wholesaler who already captured it. Neither is behaving irrationally — the arithmetic simply does not work at six sofas, which is why the wholesale layer exists.

It also explains something that confuses people about our own quotes. We can rarely improve much on a dense product bought in small volume, because the buyer was never paying a scale penalty on it in the first place. Our sea versus air calculator will show you which side of that line your product sits on faster than any conversation.

Furniture, as the worked example

Furniture is the largest category that fits the pattern: £4.19 billion a year, £3.18 a kilo, and a fragmented UK buyer base of independent retailers, hotels, offices and landlords rather than a handful of national accounts.

The tariff does not stand in the way either. Read from the UK Trade Tariff on 22 September 2026: other wooden furniture under 9403 6090 00 at 0%, metal furniture under 9403 2080 at 0%, upholstered seats with wooden frames under 9401 6100 00 at 0%, and bedding articles under 9404 9090 00 at 2%. No trade remedy naming China appeared on any of them.

So the gap is real and it is not taxed away. What stands in the way is a fire-safety regime that does not scale down either, which is the part a demand table cannot show you.

The gate that stops it being free money

Upholstered furniture sold in the UK falls under the Furniture and Furnishings (Fire) (Safety) Regulations 1988. Regulation 15 is blunt: no person shall supply any furniture or other article in respect of which any of the requirements of these Regulations is not satisfied. There is no small-importer exemption in that sentence.

Regulation 13 is the one that decides whether a small importer can really do this. It applies to any manufacturer or importer who has supplied furniture within the previous five years, and requires them to give an enforcement authority the results of any test prescribed by the Regulations, the means by which those results are attributed to a particular description or batch, and the correspondence between records and the labels or batch numbers on the goods.

Read that as a systems requirement rather than a testing cost. You are not being asked to hold a certificate; you are being asked to prove, five years later, that a specific test result belongs to a specific batch of sofas. A factory that emails you a PDF has not given you that, and our furniture sourcing guide covers what the evidence chain has to look like.

This is also why the wholesale layer is not pure rent-seeking. Part of what a furniture wholesaler is charging for is carrying that obligation, and a buyer who imports direct has taken it on personally.

How to read your own category off this

Three questions, in order. Is the category large enough in the UK that a wholesale layer exists — because if it does, somebody is already capturing a margin you might capture instead. Is it bulky, meaning under about £5 a kilo, because that is where scale stops being a preference and becomes the barrier. And does the compliance obligation scale down, or is it the same work at one pallet as at forty containers.

Where the answers are yes, yes and yes, a smaller buyer really is locked out by arithmetic rather than by competence, and consolidating demand is the way in — a shared container, a longer first order, or a group of buyers with the same specification.

Where the answer to the second question is no, be sceptical of anyone promising a transformation. On a dense product in modest volume, the freight penalty you are supposedly escaping was never large.

Here is the sentence that costs us. The categories on this page where we add the most value are also the ones where the compliance obligation is heaviest, and for a first-time buyer those two things arrive together. If you are testing whether importing works for your business at all, a dense, low-obligation product in a small quantity is a better first order even though it is the one where we save you the least. We would rather you learn the process on something forgiving than have your first container be upholstered furniture.