Is Chinese Porcelain Tile Actually Good Now? What Buyers Can Measure
Six years of buying tile out of Foshan and I still get asked the same question at least twice a month: is Chinese porcelain tile actually any good, or is it all ruined slabs and photocopied marble? The honest answer is that the country tag tells you almost nothing, and the things that do tell you something are mostly things you can measure in forty minutes on a factory floor.
I want to stay away from the usual "China has improved a lot" line, because it hides where the improvement happened and where it did not. Print quality on the good lines is now genuinely excellent. Glaze consistency between batches is still where money quietly disappears. And the gap between a serious tile factory and a trading company posing as one has widened, not closed.
Start with water absorption, not with the marketing
Everything downstream depends on what the body does with water. Porcelain is roughly below 0.5% absorption; the European classification splits it further into BIa (0.5% and under) and BIb (0.5 to 3%), then BIIa and BIIb for the semi-vitreous and porous ranges. A Chinese factory will happily quote "porcelain" for something sitting at 1.8%, because in domestic trade language the word is used loosely.
The test is boring and anyone can run it. Dry a sample at 110°C to constant weight, weigh it, boil it for two hours, soak for another two hours, reweigh, calculate the difference. I have seen declared "0.1%" turn into a measured 0.6% often enough that I now insist on being in the lab, or at minimum on getting a video where the balance displays are readable at each step. Not a PDF. A video with the sample label visible.
Floor-to-ceiling large format is where this matters. A 3200×1600mm slab at 6mm thickness with an absorption of 1.2% will bow and move after installation in a way that no adhesive specification can rescue, especially over underfloor heating.
The print is better than the marketing admits, the gloss control is not
Digital inkjet printing in Foshan caught up on resolution years ago; the 600 dpi gap people still quote in 2020-era blog posts is dead. What separates scanners at a distance is not dot resolution but three things: how well the design repeats across the batch, whether the gloss level of the printed area matches the unprinted area, and whether the factory changes glaze suppliers between orders.
The repeat issue is easy to spot if you know to look. Ask for 30 pieces of one SKU, random selection from stock, not from the sample room, and lay them on the floor in a 5×6 grid in natural light. Walk to one end and look back along the low angle. A pattern repeating every third or fourth tile is visible immediately. Some factories do this deliberately — a three-face or six-face random pack is a cost decision — and a two-face pack on a 900×1800mm slab looks like a wallpaper in a big room.
Gloss mismatch between printed and unprinted areas comes from the ink layer sitting under a different amount of protective glaze. Under a window at a low angle it reads as a faint outline around the printed vein. Under normal light nobody sees it. Your customer will see it, at 7am, with the blinds open.
Fingerprints. Polished porcelain shows them. Matt and silk finishes less so. If your market is high-end residential, ask for a full-polish versus a "polished with protective nano coating" sample and let them sit for two weeks, wipe once with a dry cloth, and see which one still shows marks.
What "calibre" actually protects you from
Calibre (the +0 or -0 suffix on the carton) is not marketing fluff, it is the factory telling you which length tolerance band the tile came out of. Tiles of the same calibre are within a defined band of each other; tiles of different calibre are not. On rectified 600×1200mm this matters at the 1mm joint you are trying to run.
Three practical rules I apply on every order:
- One calibre per room, one tone per order. Splitting a project across two production runs and getting two tones is the single most common claim I see on tile deliveries from China.
- Write the calibre band into the purchase order as an acceptance criterion, not a note. "Mixed calibre not acceptable in a single carton" is a sentence the QC team can actually act on.
- Order 3-5% overage from the same calibre batch and store it. Two years later, that box is the only thing that will save you on a repair.
My rule of thumb: if the factory cannot tell you which kiln line your order ran on, you are dealing with a trader.
Rectified edges and the 1mm joint lie
Rectified means the fired tile is cut after firing to an exact size. It does not mean the edge is square. The corner squareness tolerance on a Chinese rectified porcelain tile is typically ±0.2 to ±0.3mm, which sounds tight until you multiply it across a 3mm joint and see the lippage three rows over.
The test that has never lied to me: take six tiles, put them edge to edge on a flat surface with a 2mm spacer, and put a one-metre steel rule across the joint. Any visible gap under the rule, in either direction, is your real-world lippage. Do this in the factory, before you pay a deposit on 4000m².
Slip resistance: the R-rating is not the number your insurer wants
R9 to R13 comes from the German ramp test (DIN 51130), which is measured in oil — useful for kitchens and factories, less useful for a hotel pool deck in a warm climate. For wet areas, R values overstate safety significantly. Ask for the barefoot ramp test (DIN 51097, classes A, B, C) if the application is a pool surround, and for the pendulum test (EN 13036-4, PTV in wet conditions) if it is a public exterior.
I have seen a slab quoted as R11 with a wet PTV of 26. For a public building in Northern Europe, that is a problem, regardless of what the R class says.
The cost of a tone mismatch, in real money
Here is a case from early last year. A buyer in the Netherlands ordered 2600m² of a 600×1200mm marble-look rectified tile for a residential project. Two containers. The factory split the run across two kilns to meet a 35-day delivery promise, and the buyer accepted because the sample matched the sample room. On arrival, the tone difference between container one and container two was visible in the loading bay, before a single tile was laid.
The factory was not wrong to run two kilns — a 2600m² order on a single line is a 50-day build. What killed it was that nobody asked, in writing, whether the order would be split, and nobody wrote a tone-difference acceptance clause. In the end the factory reproduced 600m² on the original kiln line at cost; the buyer absorbed the re-shipping and lost six weeks of install.
Cost of the fix: roughly the price of the re-shipped container plus six weeks of a crew standing around. Cost of the preventive measure: one email.
Landed cost, honestly
Chinese porcelain tile is genuinely cheaper than Italian or Spanish of comparable quality on FOB price — often 30 to 50 percent — but the gap narrows once you add freight, the breakage rate in LCL, and the exchange-rate risk. A 600×1200mm 9.5mm tile weighs about 14.5kg per piece and about 20 pieces per euro pallet; by the time that is packed, palletised, and shipped to a European port, freight can be 8 to 15 percent of landed cost. If your order is under 800m², LCL breakage on tile is a real problem — I bet on 1.5 to 3 percent damage on tile LCL and it usually lands there.
FCL is better for anything above a container. 20ft takes around 800-900m² of 600×1200, 40HQ around 1800-2000m² depending on packing. Ask for the factory's own container loading photos, not the shipping agent's.
What I would do on the next order
Ask for the last three production batches, not the sample room. Demand a lab video, not a lab PDF. Take a calibre band and write it into the PO. Walk the stock randomly. If you cannot get most of that, the price you are being quoted is a price for a risk you have not priced yet.
If you want the fuller measurement stack for tile — absorption, PEI abrasion, R-value, tone and calibre — our ceramic tile buying guide goes through it step by step. And if you are buying for a project rather than retail, the FF&E buying note covers the timing side, which is where most tile projects actually break.