Fake Van Cleef: Real vs Fake Alhambra Motif, Serial and Clasp

A fake Van Cleef fails a ruler and a scale first: a Vintage Alhambra motif spans 15 mm and a 5-motif bracelet weighs 10 to 12 g. Most copies miss both numbers.

Van Cleef & Arpels publishes the motif sizes, the bracelet lengths and the prices. Auction houses publish the weights and the serial formats when they sell a piece. French law publishes what the hallmarks must be. Those three sources give the checklist, and every card on this page came from one of them. Nothing here relies on a comparison photo, because copies improve faster than photos age.

Run the four quick checks with the piece in your hand or on a video call. Then measure one motif and weigh the piece. A bracelet at 13 mm and 7 g has failed twice, and the loupe work can wait.

Vintage, Sweet, Magic and Pure Alhambra: What Changes Between Lines

The four Alhambra lines differ by motif size, and the size is the first measurement to take. Van Cleef & Arpels gives the figures in its own motif size guide. Sweet Alhambra is 9.5 mm, Vintage Alhambra 15 mm, Pure Alhambra 16 mm and Magic Alhambra 26 mm. The guide adds that each piece is handmade and the size may vary slightly. A half millimetre either way is normal. Two millimetres is not.

The Vintage line is the one copied most and the one with the clearest rule. Every motif on a Vintage piece is the same size and carries the beaded gold edge. The Magic line deliberately mixes sizes on a single necklace, so a Magic piece with three different motif diameters is correct, and a Vintage piece with three different diameters is not. Pure Alhambra drops the beads for a polished, slightly domed frame. A beaded 16 mm motif is neither line.

Stones narrow it further. The official collection page lists mother-of-pearl, onyx, carnelian, malachite, tiger’s eye, grey mother-of-pearl, enamel and diamonds across the lines, with the Byzantine line made only in yellow gold. Turquoise was used on older production and appears on circa 1970 pieces at auction. A listing for a new turquoise Vintage Alhambra bracelet at a boutique price deserves a question about when it was bought.

The history matters for one reason: the 20-motif long necklace is the original. The Van Cleef & Arpels heritage gallery dates the first piece to 1968, in creased yellow gold with beaded edges. Sotheby’s describes that first design as an opera-length necklace of 20 clover-shaped motifs. That pattern still measures 85 to 86.5 cm at auction today, so a sautoir at 80 cm or 90 cm deserves a question before the loupe comes out.

Hallmarks and the Serial Number: What Sits on the Clasp

The markings on a genuine Alhambra piece sit on the hardware at the clasp, not on the back of a motif. The RealReal locates the hallmark on the ring near the clasp. Fashionphile describes a metal plaque carrying the serial number, with the marking VCA Au750. Several resale authenticators repeat the same placement. One collecting guide claims the serial is laser-engraved on a motif back, and that claim stands alone against six sources. Treat a serial on a motif as a reason to look harder, not as proof either way.

Three marks should be present on a piece made for the French market. The French parliamentary report on precious metals sets them out. The maker’s mark, in a lozenge for manufacturers, gives traceability. The guarantee mark certifies the gold content and takes the form of an eagle’s head for gold. The fineness itself is 750 thousandths for 18 carats. The report also records that gold pieces under 3 g are exempt from the guarantee mark, which matters for a single Sweet Alhambra pendant. French customs adds that an importer’s mark is an oval rather than a lozenge. The shape of the small punch beside the eagle therefore tells you whether the piece was made in France or imported.

Pieces sold outside France follow the destination’s rules. In the United Kingdom, the Birmingham Assay Office requires a sponsor’s mark, a fineness mark such as 750, and an assay office mark on any gold article of 1 g and over. A UK-bought Alhambra bracelet should therefore carry a UK hallmark in addition to the house marks. In the United States there is no assay office; a 1982 necklace sold through 1stDibs was stamped VCA 18k with a serial and nothing else. Ask where the piece was first sold before deciding which marks are missing.

The serial number has no published format, so the auction record is the only reference. Sotheby’s catalogues give these examples, each described as signed VCA and numbered:

The pattern in that table is a tabulation of auction records, not a statement from Van Cleef & Arpels. 1970s pieces carry a plain number and a two-digit copyright year. Later pieces carry one or two letters and four to six digits, with the letter prefix changing over time. A dealer listing a 2021 hammered gold bracelet gave a serial beginning JC. None of this can be verified against the house, because Van Cleef & Arpels runs no public serial lookup. A serial that matches a real piece sold online is exactly what a copy would carry.

Weight and Chain: Why Copies Come In Light

Weight is the check that costs a counterfeiter the most to pass. A Vintage Alhambra 5-motif bracelet is 18K gold with eight prongs and a beaded frame on every motif, and five independent sales put it between 10.0 and 12.0 g. Sotheby’s sold a yellow gold and mother-of-pearl example at 11.52 g and a 17 cm white gold turquoise example at 10.00 g. CIRCA recorded 11.7 g for a carnelian bracelet. Two 1stDibs dealers recorded 11.6 g and 11.9 g. A plated brass copy at the same size weighs about 40 percent less, because brass is 8.5 g per cubic centimetre against 15 to 16 for 18K gold. The gold guide on this site explains the density test that separates the two when the weight alone is ambiguous.

The weighing procedure takes two minutes:

1. Zero a scale that reads to 0.01 g and set the closed bracelet or necklace on it.
2. Compare the reading with the band on the matching card. A 5-motif bracelet is 10 to 12 g, a 10-motif necklace 26 to 30 g, a 20-motif sautoir 41 to 50 g and a Sweet bracelet 6.7 to 7.1 g.
3. Measure the length laid flat, clasp to clasp. A 5-motif bracelet is 17 or 19 cm; a 10-motif necklace is about 42 to 43 cm; a 20-motif sautoir is 85 to 86.5 cm.
4. Divide weight by length. A 5-motif bracelet at 19 cm gives about 0.6 g per centimetre. A copy at 19 cm and 7 g gives 0.37.
5. Treat any reading outside the band as one failure, and a reading outside the band plus a wrong length as two.

A solid 18K copy can hit the weight, so the chain is the second check. Fashionphile’s authentication guide describes each link as having a pointed end and a rounded end, with a solder mark at the point where the link was closed by hand. 1stDibs repeats the same description and adds the prong rule: four rounded prongs on one face of each motif and four slightly elongated prongs on the other. Copies use identical prongs on both faces and machine-closed links with no solder point. Under a 10x loupe the difference is visible in seconds. Without a loupe it is not visible at all, which is why the weight comes first.

Mother-of-pearl is the last material check and the one a phone camera catches. Real mother-of-pearl throws green and pink light when tilted under a lamp, because the shell is built of layers that interfere with light. The shell-look resin in cheap copies is flat, uniform white with no colour shift. Onyx and carnelian copies are harder to separate by eye, so on those pieces the weight and the hardware carry the decision.

Counterfeit Van Cleef: What Customs and the Courts Show

The scale of the copying is documented in federal seizures. In April 2025, US Customs and Border Protection in El Paso seized 1,708 pieces of counterfeit Van Cleef & Arpels jewellery in one shipment from China. The manufacturer’s suggested retail value was $9.2 million; the declared value was $15,480. In June 2025, CBP officers in Louisville stopped a Hong Kong shipment bound for Tampa holding 2,193 pieces valued at $25 million. Of those, 775 were Van Cleef & Arpels copies: 490 necklaces, 205 pairs of earrings and 80 rings. In August 2025 the same port seized 7,319 pairs of fake Alhambra earrings declared as a metal necklace, with an MSRP of $30.4 million.

The court record shows the price band. Richemont, the parent of Van Cleef & Arpels, sued a New Jersey wholesaler in February 2025 over Alhambra copies priced from $40 to $700. Three cease-and-desist letters sent between 2020 and 2024 had gone unanswered. A second suit, filed in July 2025 against a New York jeweller over Cartier and Van Cleef & Arpels superfakes, ended in a consent judgment in July 2026. Richemont received $205,000 in damages and a permanent injunction. The term superfake in that filing is the point. The copies at the top of the market are solid gold with real stones, and they fail on hardware and hallmarks rather than on weight.

Those numbers set the buyer’s expectation. Fakes arrive by the thousand, cost the seller tens of dollars, and are sold at every price between $40 and the genuine article. A price that looks like a bargain against the $5,350 official figure is not a discount. It is the counterfeiter’s margin. Live-stream sales are a common channel for this stock, and the Whatnot review on this site covers what the platform’s buyer protection does and does not cover on jewellery.

Where To Get Your Van Cleef & Arpels Alhambra Authenticated

Van Cleef & Arpels itself is the only party that can check a piece against its archives. The house’s authentication service combines archival research and a technical study and produces a formal document. It takes up to six months, costs $1,730 including taxes in the United States and is requested through a boutique. The separate valuation service is offered only for pieces bought directly from the house. For a 20-motif necklace at $19,800 the fee is worth paying before a private sale; for a Sweet bracelet at $1,630 it exceeds the piece.

Resale platforms with in-house testing are the practical middle ground. Fashionphile states that it runs X-ray fluorescence on every Van Cleef & Arpels piece and issues its own certificate, which confirms the gold alloy but not the maker. eBay’s Authenticity Guarantee routes Van Cleef & Arpels items to jewellery experts at GIA before delivery and refunds in full if the piece fails. Auction houses catalogue weight, length, serial and hallmarks for every lot, which is why their records supply most of the numbers on this page. A private seller who cannot supply a weight to the gram has not looked as closely as an auction house would.

Photo authentication services fill the gap before money moves. The other signed pieces copied at this volume are covered in Cartier Love bracelet real vs fake and, for unsigned metal and stones, how to tell if a gold chain is real, fake diamond names and real vs fake pearl. LegitGrails lists Van Cleef & Arpels checks at $15 with turnaround options from 30 minutes to 12 hours. Legitly sits in the same tier: it reads your photos for the signature form, the plaque layout, the mother-of-pearl colour shift and the motif proportions, and flags what does not fit. A photo cannot weigh a bracelet or find a solder point under a loupe, so Legitly pre-screens and does not certify. The Van Cleef & Arpels hub and the wider jewellery category collect the related guides on this site. The density tests behind the weight band are in how to tell if gold is real, and the Cartier serial number check decodes the other French house whose serial buyers most often ask about.

Verdict: Two Numbers Out of Range Means Walk

A genuine Alhambra piece passes every card on this page, so one pass means little and one fail means a lot. Fakes are built to pass the checks buyers run in a photo: they carry a VCA plaque, a 750 stamp and a serial copied from a real listing. What fakes struggle to pass at the same time is a 15 mm motif, a weight inside the documented band, hand-soldered links and two different prong shapes. Each of those costs money to reproduce.

Rank the checks by that cost. A stamp costs nothing. A copied serial costs nothing. Real mother-of-pearl costs a few dollars. Solid 18K gold at the right weight costs about half the retail price, which is why superfakes exist and why they are sold at superfake prices rather than $40. Hand-closed links and asymmetric prongs cost workshop hours that a copy shop does not spend.

Two failures end it. A bracelet whose motifs measure 13 mm and whose weight reads 7 g is a copy, whatever the plaque says. A necklace at the right weight with identical prongs on both faces and no solder points has failed twice on the loupe. A piece with a serial that matches a real auction lot but the wrong clasp style for that era has failed on the one check the counterfeiter thought was safe. Treat any two failed cards as a fake and walk, whatever the box, the pouch or the story says.