The Movement Room

VS3235 vs the movement everyone confuses it with — what’s actually different

By TheRepTable · current as of August 2026

Post a Datejust caseback shot in any forum and within ten minutes somebody will tell you what’s inside it. Half the time they’ll be wrong, and they’ll be wrong in one of two specific ways. Either they’ll call a VR3235 a VS3235, which is like calling a re-bodied Corolla a Camry. Or they’ll insist the VS3235 is VS Factory’s own in-house calibre, which it has never been. The first mistake is expensive. The second is folklore, but it’s the reason the naming makes no sense to anyone who hasn’t been reading QC threads for five years. Both are worth clearing up, because as of mid-2026 the movement itself changed — and most of the listings describing that change are getting it half right.

What the genuine 3235 actually is

You can’t measure a copy without the original on the bench, so start there. Rolex introduced calibre 3235 in late 2015 in the Pearlmaster 39 and brought it to the Datejust 41 reference 126334 in 2016. It is 28.5mm across, runs 31 jewels at 28,800 vph, and carries roughly 70 hours of reserve — up from 48 on the outgoing 3136. Rolex redesigned around 90 percent of the parts and filed fourteen patents doing it.

Three of those changes matter here. The Chronergy escapement uses a skeletonised escape wheel and a nickel-phosphorus pallet fork, both paramagnetic, both cut to a geometry nothing else in the industry shares. The blue Parachrom hairspring carries a genuine Rolex overcoil. And the balance is free-sprung, regulated by Microstella nuts on the balance rim rather than by an index arm — something Rolex has been doing since long before the 3235, and the detail this whole piece eventually lands on. Certified output is −2/+2 seconds a day. That is the target. Nothing on the REP side hits all three.

The VR3235 — the movement everyone confuses it with

The VR3235 gets mistaken for the VS3235 constantly, because both appear under 3235-era Datejust and Submariner listings and both have “3235” in the name. Architecturally they are not the same class of object.

The VR3235 is a VR3135 wearing a 3235 costume. The mainplate, the gear train and the calendar work are carried over from the older 3135-derived platform, then reshaped enough to drop into a case designed around the newer movement. It keeps the 3135’s split calendar mechanism instead of the 3235’s integrated instant-jump. It runs two fewer shock absorbers than genuine spec. And critically, it uses decorative cover plates that hide the actual bridge layout underneath — so through a display back or in a QC photo, it presents a bridge geometry that isn’t structurally there. Old wine, new bottle, as one of the teardown write-ups put it.

None of that makes it a bad movement in isolation. It runs, it keeps time, and it has been in circulation long enough that parts and watchmakers know it. But it is a 3135 answering to a 3235’s name, and a buyer paying VS3235 money for one is paying for a decal.

The VS3235, and the honest limits of it

The VS3235 is a different exercise. It was drawn around the genuine 3235 rather than adapted toward it: the mainplate, the keyless works, the gear-train layout, the automatic winding assembly and the integrated instant-jump calendar all follow the genuine architecture rather than approximating it. The visible bridges are the working bridges. There are no dummy plates. Tested reserve lands at 68–72 hours, amplitude at 280–310 degrees, beat error under 0.2ms, and a VS Factory-regulated example typically runs +2 to +5 seconds a day.

That is genuinely good, and it is where honest description has to stop. The ball-bearing rotor uses seven balls in the race where the genuine part is fully packed. There is no skeletonised Chronergy escape wheel — no REP-side 3235 has one — which changes the pallet fork geometry and is precisely why genuine escapement parts do not drop in. The hairspring is not a Parachrom overcoil. And the pearlage stops at the surfaces you can see; lift a bridge on a Dandong-built movement and the hidden faces are plain, where Jinghe decorates them. None of that stops the watch working. All of it is the difference between “follows the genuine layout” and “is the genuine movement,” and the gap is wider than the spec line suggests.

So why is it called VS3235 and not DD3235?

Because both names are correct, and the community never agreed on which convention to use.

The movement is not made by VS Factory. It is made by the Dandong movement factory in Liaoning, which builds movements and sells them to watch factories. VS Factory buys it, applies its own quality control and regulation pass, and ships it as an exclusive variant — the generic Dandong 3235 goes to multiple factories and typically runs +4 to +10 seconds a day against VS Factory’s +2 to +5. Same architecture, different bin, different final regulation.

So DD3235 names the house that builds it. VS3235 names the factory that commissioned and regulates it. Neither is wrong. The reason it feels wrong is that the naming convention in this market is not a convention at all — SH3135, DD3235 and JH3235 are named after movement houses (Shanghai, Dandong, Jinghe), while VS3235 and VR3235 are named after the watch factories that deployed them. Whichever label got typed first in the review thread that broke the movement is the one that stuck. This is why you’ll see a dealer listing read “VS3235 free sprung DD3235” and assume someone’s hedging. They aren’t. That string is one movement named twice.

The 2026 change, and the argument about it

This is the part older reference material gets wrong, including earlier versions of this desk’s own record. For years the flat statement was that Dandong-built movements are not free-sprung. That was true until June 2026, when VS Factory shipped a free-sprung VS3235 in the titanium Yacht-Master 226627, and it stopped being true entirely in July 2026 when the same movement rolled across the Datejust 41 126334 range — grey dial, green ombré, blue stick, the lot. The change is specific: the regulator post above the balance wheel has been removed. Reserve is unchanged at roughly 70 hours, and stability by all accounts is unaffected.

Whether that makes it a free-sprung movement is genuinely contested, and both positions are reasonable. The case for: free-sprung means no index regulator acting on the hairspring, and VS Factory has physically removed it — the caseback view now reads correctly, which for most buyers is the entire point. The case against: on a genuine free-sprung balance, removing the index is only half the design, because the regulating function moves to weights on the balance rim. The VS Factory version has no regulating weights. Take the index off without putting the weights on and you haven’t upgraded the adjustment mechanism, you’ve deleted it, which makes fine regulation a hairspring-and-stud job rather than a screwdriver job. Some teardowns also argue the balance bridge still reads as a two-arm regulated bridge.

The competing JH3235 from Jinghe, used by QF and THBF, does carry actual regulating weights — along with a longer beat-adjust arm and a flat hairspring rather than an overcoil, so it isn’t a clean win either. The Jinghe movement is also new enough that nobody knows what it looks like after two years, where the Dandong platform is the battle-tested one. Free-sprung in appearance, argued down to regulated in function. Both descriptions are defensible, and any listing that states one without the other is selling you a headline.

How to tell what you’re actually holding

Forget the listing string. Look at the movement. Through a caseback or a QC photo, the tells run in order: is there an index arm above the balance wheel — if yes, it isn’t the 2026 batch, whatever the listing says. Is the escape wheel skeletonised — if yes, you’re looking at a genuine movement, because none of the REP-side 3235s have one. Does the calendar work read as an integrated instant-jump or a split 3135-style stack — that separates VS3235 from VR3235 faster than anything else. Are there cover plates spanning areas where the genuine movement has open bridge geometry — that’s the VR3235 tell.

Then put it on a timegrapher and take a real reading, and wind it full and see whether it clears 60 hours. A movement that does both is doing what it claims regardless of which two letters are in front of the number.

Bottom line

If a listing says 3235, the question isn’t what the number is, it’s whose it is and from when. VS3235 and DD3235 are the same movement under two naming conventions; VR3235 is an older architecture riding the number. And on any free-sprung claim made after June 2026, check the caseback for the missing index arm and treat the absent balance-rim weights as the part nobody advertises. The name on the listing has never been the thing you’re buying.

Current as of August 2026. Factory leadership shifts; this record is updated as it does.

Read next: the three movements sold as a “3235” · VS3235 and VS3285 explained · the 2026 factory map · the full index