Red dot lost zero? Find out why
By Andrew Terrebonne. Published September 22, 2026; updated October 1, 2026.
A pistol optic that loses zero has almost always moved, been moved, or is being read differently. Answer one question at a time; each answer leads to the next step, and each cause ends with the fix.
Question
What are you seeing?
Start with the mount, not the optic. It's cheaper to rule out, and it's what this site can help with.
Question
Did you witness-mark it when you installed it?
A witness mark is a paint-pen line across each screw head onto the optic or plate. It shows movement too small to feel, and tells you whether the mount moved.
Question
Does anything move by hand?
Clear the pistol first: magazine out, slide locked back, chamber checked by sight and by feel. Push the optic firmly side to side and front to back, then put a torque driver set to spec on each screw. A screw that turns before the driver clicks was loose. Don’t go past spec to find out.
Question
Which joint is moving?
A plated slide has two joints, each with its own screws and torque, and they fail for different reasons. A direct cut has one.
Question
Did it move right after a battery change?
A bottom-loading optic has to come off the pistol to change the battery, so every battery change is a re-install. Of the optics tracked here, these 4 are listed as bottom-loading:
Each optic’s page lists its battery and where it loads, where the maker publishes it.
Question
Did anything else change since it was zeroed?
Different ammunition, a different distance, a different shooter, or a different way of holding it.
Question
Does it still miss from a rest?
Shoot a slow group from a rest or a bag at the distance you zeroed at, with the dot in the middle of the window every shot.
Likely cause
The optic’s own screws
- Oil on the threads or the faces. Threadlocker doesn’t cure properly on an oily thread, and a film under the optic lets it creep. Degrease the screws, the holes and both faces.
- No threadlocker, or a spent patch. Some screws come with a dry threadlocker patch, generally good for one installation. A screw that has been in and out needs cleaning and fresh blue threadlocker (never red), or replacing.
- Torqued by feel. Each footprint has its own figure, below. “Snug” on a small screw in a thin slide is how screws back out.
- A screw that’s too long. It bottoms out before the optic is clamped, so it feels tight while the optic isn’t. See that step.
- Recoil posts not seated. An optic resting on top of a post instead of in its socket puts the recoil on the screws. See that step.
Optic screw torque, by footprint
| Footprint | Screws | Torque |
|---|---|---|
| Trijicon RMR | #6-32 | 15 in-lb |
| Trijicon RMRcc | per plate maker (plates ship their own; not the RMR's #6-32) | 18 in-lb |
| Shield RMSc | M4 × 0.7 usually; some slides M3 or #6-32 | 10–15 in-lb, per the optic or plate maker |
| Holosun K | M3, M4 or #6-32, by optic | 15 in-lb |
| Leupold DeltaPoint Pro | M4 × 0.7 | 20 in-lb (confirm current Leupold spec) |
| Docter / Noblex | M3 / #4-40 by model | 15 in-lb |
| Aimpoint ACRO | Cross-bolt clamp | per Aimpoint manual |
| Holosun 509T | per Holosun kit | 15 in-lb plate / 20 in-lb optic |
| C-More RTS2 / STS2 | per C-More kit | per C-More manual |
| Aimpoint Micro | per mount | per mount maker |
| Trijicon RCR | Trijicon capstan screws | per Trijicon's torque card |
| Aimpoint COA / A-Cut | 2 rear screws + front tongue | per Aimpoint manual |
| Holosun SCS (MOS) | per SCS kit | per Holosun manual |
| Holosun SCS (320) | per SCS kit | per Holosun manual |
| SIG Romeo1Pro | M4 × 0.7 | 28–30 in-lb |
| SIG-LOC | Screws from beneath the slide | 18 in-lb |
| SIG-LOC PRO | Screws from beneath the slide | 20 in-lb |
| SIG Romeo2 | M4 × 0.7 | 28 in-lb |
From the same data as the footprint pages. Where your optic's or plate's maker publishes its own figure, theirs wins.
Put it back on properly
A loose optic needs a re-install, not a re-tighten. Same sequence as a first install, from the fit checker’s build sheet:
- Clear the pistol. Drop the magazine, lock the slide back, look and feel the chamber, then do it again.
- Remove the cover plate and set it aside with its screws.
- Degrease the slide cut and the optic base.
- Seat the optic on the slide's recoil lugs and check it sits flat with no rock.
- Fit the optic screws with blue threadlocker and torque to spec, alternating between them.
- Let the threadlocker cure before firing (check the bottle; usually a few hours to overnight).
- Confirm zero at your working distance, then re-check torque after the first 100–200 rounds.
On a plate, the plate goes on first with its own screws and torque; the plate step has that sequence.
Likely cause
The plate, on the slide
- The wrong screws. If the pistol came with a cover plate, its screws are in the same bag as the plate screws and they are not the same length.
- Oil, no threadlocker, or torque by feel. The same three as the optic joint, with the slide’s own figure. 7 of the 44 plate systems tracked here publish a number (GLOCK MOS — standard & full size and GLOCK MOS — 10 mm / .45 frames: 13.1 in-lb (1.5 Nm); Springfield Echelon VIS: 18 in-lb maximum; Walther PDP 1.0 optic cut: approximately 2 Nm (17.7 in-lb); CZ P-10 OR plate system: 1.3 Nm (11.5 in-lb); HK VP9 / SFP9 OR plate system: 2 Nm (17.7 in-lb, converted); Staccato / 2011 optic plate system: 25 in-lb); the rest defer to the plate maker’s instructions. Your pistol’s page says which.
- It doesn’t sit flat. A burr, debris or finish in the slide pocket leaves the plate rocking on a high spot, and the screws take the recoil.
- A plate that keeps working loose however it’s fitted. Thin slides leave little thread to hold, and a steel plate holds up better than aluminum. Fewer joints is better still: a pistol page’s “No plate needed” count lists the optics that mount with no plate.
Put it back on properly
- Clear the pistol. Drop the magazine, lock the slide back, look and feel the chamber, then do it again.
- Remove the factory cover plate and set its screws aside. They are not optic screws.
- Degrease the slide pocket, both faces of the plate and the optic base. Oil under a plate lets it move.
- Seat the plate on the slide, check it sits flat with no rock, and fit its screws with blue threadlocker.
- Torque the plate to spec, bringing the screws up evenly rather than one at a time.
- Seat the optic on the plate's recoil posts, fit the optic screws with blue threadlocker, and torque to the optic spec.
- Let the threadlocker cure before firing (check the bottle; usually a few hours to overnight).
- Confirm zero at your working distance, then re-check torque after the first 100–200 rounds.
Likely cause
A mount in the rear-sight dovetail
The clamp or set screws holding the mount in the dovetail are the weak point, so check them more often than you would a plate. A mount refitted at even a slightly different angle moves the zero with it.
Slides tracked here with no optic cut, where a dovetail mount is the route:
Put it back on properly
- Clear the pistol. Drop the magazine, lock the slide back, look and feel the chamber, then do it again.
- Drift out the rear sight. Support the slide properly so you don't mar the finish.
- Degrease the dovetail and the mount's foot.
- Fit the mount, check it's square to the bore, and tighten its clamp or screws with blue threadlocker.
- Seat the optic on the mount and torque to the optic spec.
- Let the threadlocker cure before firing (check the bottle; usually a few hours to overnight).
- Confirm zero at your working distance, then re-check torque after the first 100–200 rounds.
Likely cause
A screw that’s too long
How it usually happens: the screws in an optic’s box are cut to mount it straight onto a slide. A plate changes the stack, which is why the plate maker supplies the optic screws for a plated build. Screws from the wrong bag, or one set reused on a different stack, are the usual culprits.
Measure, don’t guess. This site doesn’t print screw lengths, because they depend on your exact stack. The fit checker’s build sheet has a screw-length check: enter two caliper readings (the stack the screw passes through, and the usable depth of the tapped hole) and it gives you the longest safe screw.
Likely cause
A screw that won’t tighten
Stop turning it. A screw that spins is either stripped or cross-threaded, and more turns make both worse.
- The wrong thread. The tapped thread is in the slide or the plate, not the optic, so screws from the optic’s box can be the wrong thread for your slide. Start each one by hand before using a driver.
- GLOCK Slimline MOS (43X / 48). GLOCK says the standard Slimline MOS cut takes M4 screws; SIG's ROMEOZero-ELITE manual gives them as M4 × 0.7 for the 43X MOS.
- GLOCK Slimline MOS-K. Tapped M3, not the M4 of the earlier Slimline MOS. The two slim GLOCK cuts look alike but take different screws; mixing them up is the most common cross-threading mistake on slim GLOCKs.
- Docter / Noblex optics. Different makers on this footprint ship M3 and #4-40, so check what’s in your box.
- A stripped hole. In a plate, replace the plate; they’re cheap. In the slide, it’s gunsmith or manufacturer work. Don’t force in a bigger screw.
Every pistol page gives the slide’s thread where the maker taps it, and every build sheet names the thread for each joint.
Likely cause
The optic won’t sit flat
An optic should drop onto its recoil posts and sit flat before you turn a screw. Never use the screws to pull it down, or the recoil goes through the screws from the first shot.
- A footprint that looks right and isn’t. Looks like a small RMR but isn't one. Needs its own cut or plate; won't sit on an RMR or RMSc pattern. Trijicon RMRcc.
- Posts the cut doesn’t have. RMSc screw spacing with the two rear recoil lugs left off. So a K optic fits an RMSc cut only where that cut has no rear lugs; on a four-lug cut such as the original GLOCK Slimline MOS it needs a plate. Whether an RMSc-pattern optic goes on a K cut the other way round is up to each slide's maker; each pistol's page says what that maker lists. Holosun packs different screw pairs with different K models, so check what came in the box. Shield RMSc vs Holosun K sets out which goes on which.
- Something in the way. A burr, debris, excess finish or a proud screw head in the pocket. Clean it and try again dry.
- Pins or a plate not fully seated. If the slide locates the optic with pins or a plate, seat those first.
Likely cause
The battery change was a re-install
Taking the optic off undoes the threadlocker, the torque and the witness marks. Treat every battery change as a re-install (degrease, fresh blue threadlocker, torque to spec, new marks) and check zero before you trust it.
If you’re still choosing an optic, a top- or side-loading battery avoids this. Each optic’s page says where its battery goes, when the maker publishes it.
Likely cause
Something else changed
- Ammunition. A different load can land somewhere else. Zero with what you carry.
- Distance. A dot zeroed at one distance lands differently at another. Check zero at the original distance before deciding it moved.
- Shooter or grip. A group that moves with a new shooter or grip is usually the hold. See that step.
Likely cause
It’s the hold, not the hardware
If it groups on zero from a rest, the optic and mount are fine. What’s left is grip pressure, trigger press, and where the dot sits in the window. Keep it centered: head-position errors show most at the edge. Work outward from the rest, and leave the adjusters alone.
Likely cause
The optic itself
If the marks haven’t moved, nothing turns by hand, the ammunition and distance are the same, and it still misses from a rest, the problem is inside the optic. Contact the maker and stop re-zeroing it until then. Tell them what you’ve ruled out; they’ll ask.