← All Guides

Is that SKX007
genuine, modded, or built from parts?

The SKX007 spent twenty years as the cheap watch nobody would bother faking. Then Seiko discontinued it, prices tripled, a mod-parts industry grew around its case — and now the question “is my SKX real?” fills forums weekly. The answer is rarely a simple counterfeit. It is usually something more interesting.

Last reviewed July 2026 · Independent guide · Not affiliated with Seiko SKX007

The 20-second answer

Three different things get sold as an SKX007: genuine watches, genuine watches wearing aftermarket parts, and complete builds assembled from mod-industry components around a real Seiko-family movement — often an NH35 rather than the correct 7S26. The dial is where they betray themselves: on a genuine dial the word Automatic is printed slightly wider than the word Seiko above it and in a bold weight, while copies print both at matching width in a normal weight. The bezel pip is the second tell — genuine inserts carry a recessed lume dot under a window, copies a raised painted drop — and the hands are the third: Seiko stamps the arrow into the metal itself, copies merely paint an arrow shape onto a plain hand. And remember Seiko’s own warning covers modified watches as well as counterfeits: an SKX that hacks or hand-winds is running a swapped movement, whatever the listing says.

GENUINE SKX DIALSEIKOAUTOMATIC — WIDER, BOLDAutomatic wider than Seiko; bold weight; crisp edgesApplied, framed indices with even lumeDIVER’S 200m in correct weightPrint sits crisp under magnificationAFTERMARKET / COPY DIALSEIKOAUTOMATIC — SAME WIDTHEqual widths, normal weight — the copy’s signatureFlat or glued markers, uneven lumeFont weight uniformly thinHalo or fuzz around characters
Fig. 1 — The font tell. Forum authentication threads resolve on this detail over and over: genuine SKX dials print Automatic slightly wider than the Seiko wordmark above it, in a distinctly bold weight. Aftermarket and counterfeit dials print both words at matching width in a normal weight. It survives even decent photography, which makes it the best remote check available.

The six checks, in the order they pay off

Run these against good photographs before money moves. Each one catches a different class of problem.

Dial font and weight

Genuine: Automatic printed slightly wider than Seiko, in a bold weight. Copies: equal widths, normal weight. This single comparison settles a remarkable share of cases.

Bezel lume pip

Genuine inserts have a recessed pip — a lume dot sunk beneath a small window at 12. A raised, painted, domed drop of lume is aftermarket; the recessed design is rarely copied.

Hand construction

On genuine hands the arrow is stamped into the metal and lume fills the shape. On copies the metal is a plain paddle and only the paint makes the arrow.

Crown behaviour

A real SKX runs a 7S26: no hacking, no hand-winding, crown at 4. If the seconds hand stops or the crown winds, the movement has been swapped — commonly for an NH35/NH36.

Caseback code and serial

SKX007 and 009 casebacks read 7S26-0020 (the 013 reads 7S26-0030), with a serial whose first two digits give year and month. Cross-check the implied date against the movement revision and lume aging.

J vs K, correctly

The J (Japan-made) dial carries 21 JEWELS and MADE IN JAPAN text; the K does not. A K missing those lines is not fake — it is a K. Panic over this difference sells a lot of genuine watches cheap.

Why the SKX became worth faking

For most of its 1996–2019 production run, the SKX007 was the answer to “what dive watch should I buy first?” — an ISO-rated 200-metre diver with Seiko’s indestructible 7S26 inside, sold new for pocket money. Nobody counterfeits a plentiful cheap watch. Then Seiko discontinued the line, the enthusiast press canonised it, and prices for clean examples multiplied. Scarcity plus sentiment is exactly the recipe that attracts imitation.

What makes the SKX case unusual is that the imitation industry was already in place, wearing a friendly name. The Seiko modding scene — a genuinely creative, legitimate hobby — had spent years commissioning aftermarket SKX-compatible cases, dials, hands, inserts, crystals and chapter rings. Every part of an SKX exists as a reproduction because modders wanted options. The moment originals became valuable, the same catalogue that served hobbyists began serving deception: complete watches assembled from aftermarket parts around a cheap NH35, photographed well, and listed as “Seiko SKX007”.

Seiko itself has acknowledged the problem from both ends, issuing warnings covering counterfeit watches and modified watches sold as original — a distinction that matters, because the second category is far larger than the first. A pure counterfeit SKX exists but is rare; a listing photograph of aftermarket parts on a genuine case, or of a ground-up build, is Tuesday.

This is why authentication here is really three questions, not one: is any of it Seiko, is all of it Seiko, and did all of it start life together? The checks below are ordered to answer them in that sequence.

GENUINE SKX DIALSEIKOAUTOMATIC — WIDER, BOLDAutomatic wider than Seiko; bold weight; crisp edgesApplied, framed indices with even lumeDIVER’S 200m in correct weightPrint sits crisp under magnificationAFTERMARKET / COPY DIALSEIKOAUTOMATIC — SAME WIDTHEqual widths, normal weight — the copy’s signatureFlat or glued markers, uneven lumeFont weight uniformly thinHalo or fuzz around characters
Fig. 2 — The font tell. Forum authentication threads resolve on this detail over and over: genuine SKX dials print Automatic slightly wider than the Seiko wordmark above it, in a distinctly bold weight. Aftermarket and counterfeit dials print both words at matching width in a normal weight. It survives even decent photography, which makes it the best remote check available.

The dial: where copies still lose

Aftermarket SKX dials have become good — good enough that colour, layout and general finish no longer separate them. Two details still do, and both survive photographs. The first is the font tell: on a factory dial, the word Automatic is printed slightly wider than the Seiko wordmark above it, in a noticeably bold weight. Copies overwhelmingly print the two words at matching width, in a normal weight. Once seen, the difference is hard to unsee, and forum authentications turn on it constantly.

The second is construction rather than printing. Genuine SKX indices are applied — discrete framed markers with lume inside a surround — and the lume is machine-applied, even and contained. Budget aftermarket dials simulate the look with printing or glued blocks, and their hand-applied markers betray themselves in alignment: individual indices sitting a hair off their positions, or lume that varies plot to plot.

Here the SKX offers its best joke, with a real lesson inside. Genuine examples are notorious for misaligned chapter rings and slightly rotated bezel inserts — enthusiasts quip that perfect alignment is the suspicious outcome. The lesson is not that misalignment proves authenticity; it is that alignment tests the wrong thing on this model. Seiko’s tolerances were loose at this price. The copies fail on font metrics and marker construction, not on the crookedness the factory itself never mastered.

Photograph requests that settle it remotely: a straight-on dial shot in even light for the font comparison, and a raking-angle shot for marker construction and lume texture. A seller who cannot supply both is making a choice.

WHERE AN SKX GIVES ITSELF AWAY1Recessed bezel pipLume sunk under a window — raised paint is aftermarket2Stamped arrow handsArrow formed in the metal, not painted on3Crown at 4, no hackStops or winds — movement swapped47S26-0020 casebackSerial’s first two digits date it
Fig. 3 — Four physical checks beyond the dial. Each targets a different failure: the pip catches replaced inserts, the hands catch the commonest aftermarket swap, the crown catches movement replacement instantly, and the caseback anchors the whole watch in time so everything else can be checked for consistency against it.

Bezel, hands, crystal: the aftermarket’s fingerprints

The bezel insert’s lume pip is the most reliable single hardware tell. Factory inserts carry a recessed pip: the lume dot sits sunk beneath a little window at the 12 position, flush with the insert surface. The aftermarket — and every counterfeit yet seen — almost universally uses a raised, painted, slightly domed drop of lume instead. Run a fingernail across it in person; in photos, the raised pip catches light as a bump where the genuine article shows a flat ring around a sunken dot.

Hands are next, and the tell is structural. Seiko’s hour hand carries its arrow in the metal itself — the shape is stamped, and lume fills the stamped form, leaving a metal border defining the arrow. Aftermarket hands are typically a plain paddle onto which an arrow of lume is painted: the shape exists only in paint, with no stamped border. Under any magnification the difference is unmistakable, and it matters because hands are the single most commonly swapped part on modded examples.

The crystal completes the set. Factory SKX crystals are Seiko’s Hardlex mineral glass, slightly domed, with a characteristic warm distortion at extreme angles. The most popular mod on earth is replacing it with flat or double-domed sapphire — frequently with a blue anti-reflective coating that flashes in photographs. A blue-flashing, perfectly scratch-free crystal on a “never opened, all original” SKX is a listing contradicting itself.

None of these parts being aftermarket makes a watch worthless — the mod scene exists because these upgrades are genuinely nice. Every one of them ends the word “original”, and with it the collector price.

The movement question: one crown-pull

Every genuine SKX007 left the factory with a 7S26 — a movement that does not hack and does not hand-wind. That absence is the fastest authentication tool on the model: pull the crown fully out, and if the seconds hand stops, or if the crown winds the mainspring in first position, the movement is not the original. No exceptions, no ambiguity, thirty seconds, no tools.

What is inside instead is usually an NH35 or NH36 — the modern successors that drop straight into the case and add exactly those two features. As an upgrade for an owner, it is rational and popular. As an undisclosed fact in a sale, it is the difference between an original watch and a serviced composite, and on today’s SKX market that difference is real money. The franken builds go a step further: an NH35 is the cheapest Seiko-family movement to buy loose, so ground-up aftermarket builds almost always carry one — which means the crown test catches most complete fakes too, as a side effect.

For examples that pass, the 7S26’s revision letter offers a second layer: A ran to roughly 2006, B to 2011, C afterwards. Read it against the caseback serial’s year digit. A 1998-serial case carrying a C-revision movement has had its heart replaced at some point — common, forgivable, but not “all original, one owner”.

The full movement story — what the 7S26 is, why its absences were deliberate, and how to live with them — has its own page; for purchase purposes the crown-pull is ninety percent of the value.

GENUINE, MODDED, OR FRANKEN?Does the dial pass the font-width and weight test?Aftermarket dial — modded or builtDoes the crown refuse to hack and hand-wind?Movement swapped — not originalDo pip, hands and insert match factory patterns?Parts replaced — price as moddedDoes the serial date agree with movement revision and condition?Story does not add up — walkConsistent throughout → most likely genuine
Fig. 4 — Three verdicts, three prices. A genuine-original SKX carries the collector premium. A genuine-but-modded one is worth parts and honesty. A ground-up build from the mod industry — case, dial, hands, insert all aftermarket around a bare NH35 — is a new watch cosplaying as a discontinued one, and should cost like the former.

J, K, and the variants that get called fake

More genuine SKXs are wrongly condemned than fakes are wrongly trusted, and the J/K split causes most of it. Japan-assembled J models print 21 JEWELS and MADE IN JAPAN on the dial; overseas-assembled K models omit those lines. Both are factory variants. A K is not a J with text missing — it is a K, and calling it fake in a forum thread has talked countless sellers into discounting genuine watches.

The bilingual day wheel works the same way. Seiko fitted language pairs by market — English with kanji for Japan, with Arabic for Middle Eastern distribution, with Spanish elsewhere — so an unexpected second language is evidence of where the watch was sold, not of forgery. It only becomes a red flag when it contradicts the rest of the story: a J-dial watch wearing an Arabic wheel and a serial that dates to a K-era batch is a consistency failure, and consistency is the test that matters.

That is the closing principle for the whole model: almost every part of an SKX exists legitimately in several versions, so no single part proves anything. What convicts a watch is disagreement — a dial variant, movement revision, day wheel, serial date and wear pattern that cannot all be true of one watch at once. Genuine watches, even modded ones, tell one coherent story. Builds tell several, simultaneously.

READING AN SKX CASEBACK7S26Movement codeThe calibre the case was built for — 0020 body is the 007/0090020Case code0020 = SKX007/009 · 0030 = SKX013 midsizeBelow it, the serial’s first digit is the production year, the second the month — cross-check everything against this date.
Fig. 5 — The caseback is the anchor. Movement-case code plus a date-encoding serial makes the SKX unusually checkable: the dial variant, movement revision letter, day-wheel languages and even lume aging all have to be plausible for the year the serial declares. Frankens assembled from mixed-era parts fail this consistency test far more often than they fail any single-part inspection.
MarkingWhat it isFake-panic risk
SKX007JJapan-made · 21 JEWELS + MADE IN JAPAN on dialK buyers think their dial is missing text
SKX007KOverseas-made · shorter dial textWrongly condemned daily; it is genuine
SKX009Navy/red “Pepsi” sibling, same 0020 caseInsert swaps between 007/009 are invisible
SKX01338mm midsize, caseback 7S26-0030Sometimes sold as a “rare small 007”
Day wheelBilingual — language pair follows marketKanji or Arabic wheels read as “wrong” but match origin

Red flags on an SKX listing

Each of these has ended real purchases — or should have.

Equal-width, normal-weight dial text

The copy signature. Genuine dials print Automatic wider than Seiko, in bold. This one detail resolves more SKX threads than any other.

Raised painted bezel pip

Factory inserts carry a recessed lume dot under a window. A domed drop of paint means an aftermarket insert — and asks what else was replaced.

It hacks or hand-winds

Then it is not running its 7S26. Movement swapped — fine if disclosed and priced, disqualifying if the listing says all-original.

“All original” with a blue-flashing sapphire

Factory Hardlex does not flash blue. The listing is contradicting itself on the easiest part to check.

Scan your SKX before you pay for it

Photograph the dial straight-on, the bezel pip and the caseback. An instant read runs the font, pip and consistency checks before you commit.

Scan Your Watch

For high-value purchases, we recommend pairing your AI scan with an in-person inspection by a certified watchmaker — in-person inspection remains the gold standard.

Frequently asked questions

How can I tell if a Seiko SKX007 is fake?

Start with the dial: genuine examples print the word Automatic slightly wider than the Seiko wordmark, in a bold weight, while copies print both at equal width in a normal weight. Then check the bezel pip — recessed under a window on genuine inserts, a raised painted drop on copies — and the hands, where Seiko stamps the arrow into the metal rather than painting it on. Finish with the crown: a real SKX neither hacks nor hand-winds.

Are most fake SKX007s actually counterfeits?

No — and this is the key to the whole problem. Pure counterfeits exist but are rare. Far more common are genuine watches wearing aftermarket mod parts, and complete builds assembled from the mod industry's catalogue around a cheap NH35 movement. Seiko's own public warning covers both counterfeit and modified watches sold as original, because the modified category is the bigger one.

My SKX doesn't say 21 Jewels on the dial. Is it fake?

Almost certainly not — it is a K variant. Japan-assembled SKX007J models print 21 JEWELS and MADE IN JAPAN on the dial; overseas-assembled K models omit those lines by design. Both are factory-genuine. More genuine SKXs get wrongly condemned over the J/K difference than fakes get caught by it.

My SKX seconds hand stops when I pull the crown. Good sign?

The opposite — it means the movement has been swapped. The factory 7S26 does not hack and does not hand-wind; those absences are the specification. An SKX that hacks is running an NH35, NH36 or 4R-family replacement. That is a popular, functional upgrade, but the watch is no longer original, and it should not be priced as if it were.

What should the SKX007 caseback say?

7S26-0020 — the movement-case code shared with the SKX009 (the midsize SKX013 reads 7S26-0030) — plus a serial number whose first digit encodes the production year and second the month. That date anchors every other check: dial variant, movement revision letter, day-wheel languages and honest wear all have to be plausible for the serial's year.

Is a modded SKX worth buying?

Often, at the right price. Sapphire crystals, better inserts and NH36 swaps are genuine improvements, and the mod scene is a legitimate hobby, not a scam. The rule is simply that mods end originality: a modded example should cost meaningfully less than an untouched one, and every replaced part should be disclosed. The watch to avoid is the ground-up aftermarket build priced as a discontinued original.

Why are perfectly aligned SKXs jokingly called suspicious?

Because factory SKXs were notorious for slightly misaligned chapter rings and bezel inserts — Seiko's tolerances at this price were loose, and enthusiasts joke that flawless alignment is the anomaly. The serious point underneath: alignment is the wrong test for this model. Copies fail on font metrics, marker construction and the bezel pip, not on the crookedness the factory itself shipped.

Keep reading

Other brands