Search for a casement window and you will meet the phrase multi-point locking within about four clicks. It appears in feature lists, in quotation spreadsheets, in showroom scripts. It is almost always a number: four-point, six-point, eight-point. And it is almost always presented as though the number were a property of the window, the way the glass thickness is a property of the window.
It is not. The number of locking points on a casement sash is an output. It is what falls out when you divide the size of your sash by a spacing limit that almost nobody prints. Change the sash and the same lock stops being adequate. Change the spacing limit and the same sash needs a different number of points. The count is the answer to a question, and the question is over what length.
This article is not about which lock to buy. It is about the one number that decides everything and is missing from nearly every page that uses the phrase.
Before any arithmetic, the phrase has to be taken apart, because four different claims wear it and only one of them can be written onto a purchase order.
| What the speaker means | What it actually is | Can you write it into an order? | What it does not tell you |
|---|---|---|---|
| A count of engagement points | A number, usually counted on one configuration of one sash | Only with the sash dimensions printed next to it | How far apart those points are, which is the thing that matters |
| A mechanism type | An espagnolette drive rod with mushroom cams, or a shootbolt system driving bolts into head and sill | Yes, and you should, because the two take different keeps and different replacement parts | How many points that mechanism carries on your sash |
| A perimeter geometry | Locking points distributed around all four sides via corner drives, rather than only along the handle stile | Yes, and it is more informative than a count | Whether the hinge side carries points, or only hinges |
| A security claim | An implication that more points means harder to force | No. A security rating is issued to a tested window, not to a count | Anything at all, until you have a standard number, a version, and a certificate number |
The first row is the one that generates almost all of the confusion, because a count is the easiest thing in the world to put in a column and compare. Numbers sort. Sorting is not the same as knowing.
Here is the part that rarely makes it onto a product page. A clause in a Chinese national standard for plastic doors and windows, reproduced on more than one industry site and quoted in window-hardware buying guides, does not say "use four locking points". It says three other things:
| Dimension | Threshold | What it triggers | Who owns that number |
|---|---|---|---|
| Sash width | Greater than 650 mm | Top and bottom locking points must be added | Your opening. The factory does not choose your width |
| Sash height | Greater than 1000 mm | No fewer than three locking points on the handle side and no fewer than three on the hinge side | Your opening, again |
| Spacing between adjacent points | Not more than 700 mm | This is the clause that actually generates the count | Nobody. It is the number most often left out of the quotation |
Read those three rows in order and the whole thing inverts. The count is not chosen. It is what is left after you have obeyed the spacing limit. Which means the question to ask a supplier is not "how many points"; it is "what is the largest gap between two points on the tallest sash in my order".
The following table is entirely ours. Take a sash of a given height. Put the first and last locking point 150 millimetres from each corner — a Chinese window company's technical article puts that end distance somewhere between 110 and 250 millimetres, and 150 sits inside that range. Then ask how many equal gaps you need so that no gap exceeds the spacing limit. The answer is the minimum number of points on that side.
We run it twice, because two published figures exist for the spacing limit and they are not the same number. The 700 millimetre figure comes from the standard clause above. A separate Chinese trade encyclopaedia entry recommends a spacing of no more than 600 millimetres, with the edge point within 150 millimetres of the corner.
| Sash height | Minimum points per side at 700 mm limit | Resulting gap | Minimum points per side at 600 mm limit | Resulting gap |
|---|---|---|---|---|
| 800 mm | 2 | 500 mm | 2 | 500 mm |
| 1000 mm | 2 | 700 mm, exactly at the limit | 3 | 350 mm |
| 1200 mm | 3 | 450 mm | 3 | 450 mm |
| 1500 mm | 3 | 600 mm | 3 | 600 mm, exactly at the limit |
| 1800 mm | 4 | 500 mm | 4 | 500 mm |
| 2100 mm | 4 | 600 mm | 4 | 600 mm, exactly at the limit |
| 2400 mm | 4 | 700 mm, exactly at the limit | 5 | 525 mm |
| 2700 mm | 5 | 600 mm | 5 | 600 mm, exactly at the limit |
Two rows in that table are worth more than the rest. At a sash height of 1000 millimetres, one published limit says two points per side and the other says three. At 2400 millimetres, one says four and the other says five. Same sash, same corner allowance, two different answers, and the difference is entirely in which spacing figure somebody decided to work to.
Take a sash the size of a nominal 28 x 80 entry door opening, that is 711 by 2032 millimetres. Its width is over 650, so the top and bottom points are triggered. Its height is over 1000, so both long sides need three points at minimum. Run the spacing: 2032 minus two corner allowances of 150 leaves 1732, and 1732 divided by 700 is a little over 2.4, so you need three gaps and therefore four points on that side. Under the stricter 600 millimetre figure, the same arithmetic with the end points set closer to the corners gives five. That sash does not have "a" number of locking points. It has a number per side, decided by a spacing limit, and the two published limits disagree.
Now do the division the other way round, which is the direction a buyer actually has to do it. Somebody offers eight locking points. Divide the perimeter of the sash by eight. That is the average gap. This is our arithmetic, using a deliberately generous simplifying assumption — that the points are spread evenly around all four sides. Real hardware clusters them on the long sides, which makes the real gaps on those sides larger, not smaller, so this version flatters the lock.
| Sash, width by height | Perimeter | Average gap for eight points | Against 700 mm | Against 600 mm |
|---|---|---|---|---|
| 600 x 900 mm | 3000 mm | 375 mm | Passes with room | Passes with room |
| 700 x 1500 mm | 4400 mm | 550 mm | Passes | Passes |
| 900 x 1500 mm | 4800 mm | 600 mm | Passes | Exactly at the limit |
| 900 x 1800 mm | 5400 mm | 675 mm | Passes, narrowly | Fails |
| 900 x 2100 mm | 6000 mm | 750 mm | Fails | Fails |
| 1000 x 2400 mm | 6800 mm | 850 mm | Fails | Fails |
| 1200 x 2400 mm | 7200 mm | 900 mm | Fails | Fails |
The same eight-point lock, on the same product family, is comfortably adequate on a 700 by 1500 sash and short of the mark on a 900 by 2100 sash. Nothing about the lock changed. Only the denominator changed. And the denominator is the one thing that never appears in the column where the eight is printed.
A UK hardware retailer listing an espagnolette multipoint mechanism offers the drive rod in four stock lengths: 380, 580, 780 and 980 millimetres. The standard clause quoted earlier changes its requirement at 1000 millimetres of sash height.
Those two numbers are about twenty millimetres apart. We are not claiming the standard was written around a stock part, and we have no document that says so. We are pointing out that a very common stock component stops at almost exactly the height at which the rule starts asking for more, and that if your sash is taller than 980 millimetres, the mechanism is being extended, split, or supplemented rather than simply supplied. That is a fair question to ask at quotation stage and almost nobody asks it.
The same retailer's page gives the dimensions that actually govern whether you can ｒｅｐｌａｃｅ the thing later: a backset of 20 or 25 millimetres, a rail width of 16 millimetres with a slim 14 millimetre variant, a mushroom cam height given as 7.7 millimetres, a gearbox 60 millimetres wide and 6 millimetres thick. Another UK hardware page adds the dimension the industry genuinely did standardise: the distance between the three holes in the side of the gearbox, spindle hole in the middle with a handle screw either side, is 43 millimetres, and it is described as the same across essentially all espagnolette rods.
This is the section that changes how a count reads. A hardware manufacturer's published maintenance manual for its security espagnolette states it plainly: each locking point comprises one adjustable compression cam and one fixed security cam. The compression cam is eccentric and continuously adjustable, rotated with a 3 mm hex key through up to 360 degrees to find the optimal compression point. And then the sentence that matters most: compression must be even at all compression points along the sash.
So an eight-point lock is not eight things. It is sixteen cams, eight of which have to be individually set with a 3 mm key, and the setting has to come out equal. Every point you add is another place where the compression can be wrong, and a lock with one badly set point out of eight is a lock with a draught at that point and a handle that is fighting the other seven.
| The two halves of one point | Its job | Who sets it, and when | What goes wrong |
|---|---|---|---|
| Compression cam, eccentric, adjustable | Pulls the sash onto the gasket by a chosen amount | At the factory, then again on site after glazing, then again seasonally | Uneven between points gives a draught at one place and a stiff handle everywhere |
| Security cam, fixed | Resists the sash being levered away from the frame | Set by the geometry of the keep, not by adjustment | If the keep is misplaced, no amount of cam adjustment will fix the engagement |
| The keep, or striker plate, on the frame | The other half of both cams | On site, and it is a separate part with its own part number | A worn or misplaced keep is the usual cause of a stiff handle |
A UK hardware retailer says the same thing from the spare-parts side: when a handle turns but the window will not lock, the fault is usually the locking mechanism inside the edge of the sash, not the handle; and a stiff handle is often caused by worn or misaligned keeps on the frame. A repair guide adds that after a sash has dropped under gravity the cams strike the edge of the keep rather than sliding into it, that you can see where by looking for scrape marks on the keeps, and that moving a keep by one or two millimetres is often the whole repair.
There is a measurement more important than the count, and it is almost never quoted. It is the lap, or engagement: how far the locking point overlaps the keep once the handle is turned home.
One Chinese window company's technical article gives a working lap of 2.5 to 4 millimetres, and explains the two failure directions: too small and the point can pull out under wind load, too large and the window becomes hard to close and the gasket is over-compressed. An Australian-facing supplier's product page gives a different-sounding figure for a named European hardware system: a mushroom head engagement depth of 12 millimetres and a single-point resistance quoted at 600 newtons or more.
We are not going to reconcile those two numbers. They may be measuring different things, one may be a lap and the other a penetration depth, and we have no document that translates between them. The point of putting them side by side is that a buyer who asks "what is the engagement" will get two different answers from two different suppliers and will not be able to compare them, which is precisely why the question has to be asked with the measurement named.
A multipoint lock is a row of outputs driven by a single input. One handle, one spindle, one gearbox, one drive rod, and then however many cams are riveted along it. The count describes the output side. The reliability lives on the input side.
The repair literature is consistent on where it fails. A UK spare-parts specialist writes that when the handle turns freely but the window will not lock or open, the gearbox has usually failed, and that the gearbox is the most common point of mechanical failure because it houses the spindle gears and internal springs. A hardware retailer's listing describes endurance testing to a stated number of operations for the mechanism as a whole. A separate installation work instruction gives acceptance figures for the finished window: a locking force measured between 100 and 150 newtons, and an opening force of no more than 50 newtons.
A repair guide gives the field version of the same idea, and it is the sentence to remember: a correctly aligned lock should require no force to engage the handle, and if you have to pull the sash inward to turn it, the keeps are too tight. Those two statements — a measured locking force in newtons, and no force to engage — are different tests of different things, and we are not going to pretend otherwise. Read together they say this: the handle effort is a diagnostic, and a handle that needs muscle is telling you the geometry is wrong, not that the lock is strong.
A casement sash has four sides. The multipoint mechanism runs down the handle stile and, if the system has corner drives, round onto the head and sill and possibly down the hinge stile. The hinge side is where the weight is carried, and it is also the side an intruder goes for on an outward-opening window, because the hinges are on the outside.
The component that answers this is called a hinge bolt, or dog bolt, and a hardware supplier's product page describes it with unusual candour. It locates on the opposite side to the security locking system. It aids compression on the seals by always locating in the same position. It aids the pressure on the hinges. It prevents the sash from being crowbarred open. It fits into the Eurogroove, and the keep is sold separately. And the recommendation on that page is two bolts and keeps per window.
Two things follow. First, a dog bolt is a locking point that does compression work, which means the total number of places holding the sash onto its gasket is not the espagnolette count. Second, it is sold with the keep separately, which means the frame half of a security component is a line item that can be quietly left off an order.
A UK glazing company's security page describes the attack the bolt answers: on a leaf with exposed hinges, punch out the hinge pins and lever the leaf out of the frame on the lock side, and the answer is concealed hinges or hinge-side dog bolts that engage into the frame independently of the hinge pins. The same page gives its minimum window specification as multipoint locking with at least three engagement points per casement and four per tall sash, plus stainless friction-stay hinges with a hinge-side anti-lift bolt.
If you want to know why nobody can tell you the right number of locking points, this table is the reason. These are all published, all findable, and all different. Some key the count to height, some to width, one to risk rather than size.
| Published source, described not named | What it keys the count to | The numbers it gives | What it does not say |
|---|---|---|---|
| Assembly clause of a Chinese national standard for plastic doors and windows, quoted on industry pages | Both width and height | Width over 650 mm adds top and bottom points; height over 1000 mm needs three per side; spacing no more than 700 mm | Whether the same applies to aluminium, which is a different standard family |
| A Chinese window company's configuration article | Sash height | Up to 800 mm: two. 800 to 1200 mm: three. Over 1200 mm: add a fourth. End points 110 to 250 mm from the edge. Lap 2.5 to 4 mm | What happens above 1500 mm, where its own advice stops |
| An Australian-facing supplier's product page | Sash height, for a tilt-and-turn range | Under 1200 mm: three. 1200 to 1800 mm: five. Over 1800 mm: seven | Whether those are counted on one stile or around the perimeter |
| A Chinese hardware installation work instruction | Sash width, not height | Width up to 1200 mm: two points. Over 1200 mm: three, evenly spaced. Locking force 100 to 150 N, opening force no more than 50 N | Any spacing limit at all |
| A UK glazing company's security specification page | Casement versus tall sash, without a dimension | At least three engagement points per casement, four per tall sash | What counts as tall |
| A UK police force crime-prevention page | Size and risk together | At least one lock per opening frame; two locks opposite each hinge where the casement is bigger than one metre or in a higher-crime area | Anything about spacing, engagement, or force |
The thresholds in that table appear at 650, 800, 1000, 1200, 1800 and one metre. The counts range from two to seven. Two documents key the count to width, three to height, two to neither. There is no single correct number to quote, and there never was. There is only the number that belongs to your sash under the rule your project has to satisfy.
The buying instinct is that a bigger number is better. Two published sources say otherwise, and they say it from opposite directions.
The Chinese window company's article warns directly that surplus locking points increase engagement resistance, make the handle stiff and labouring, and that incorrect adjustment easily causes the points to sit out of position. It caps its own advice: ordinary domestic windows do not need more than four groups. A separate Chinese trade encyclopaedia entry puts the mechanism plainly — the locking points have to form a closed load ring with the hinges, and adding points on its own does not raise the strength of the whole.
That last sentence is the one to take to a supplier. The load path runs from the sash, through the cams, into the keeps, into the frame, into the fixings, and into the wall. A keep screwed into a thin reinforcement is the weak link in a chain with eight links, and adding a ninth link does not strengthen it. A UK glazing company says the same in specification language: reinforced steel keeps mortised into the frame, not screwed into the reinforcement.
The hinges carry the weight, and their number scales too. A Chinese hardware installation work instruction positions hinge centres at one tenth of the frame height from the top and bottom edges, not less than 100 millimetres and not more than 200 millimetres, and adds one more hinge when the frame height reaches 1800 millimetres. It also specifies the torque for the fixing screws at 4 to 6 newton metres, tightened in diagonal sequence rather than one at a time, and a coaxiality tolerance between upper and lower hinges of no more than 0.2 millimetres. Those are the numbers that decide whether the locking points ever line up with their keeps.
Chinese reporting after a recent typhoon, quoting a glass industry technical chief, sets out the two failure directions with unusual clarity. An inward-opening window that is not locked can be blown open by positive pressure. An outward-opening window that is not locked can be sucked open by negative pressure on the outside. And the sentence that should be on every handover sheet: a window whose locking points have not been rotated home counts as not closed.
The same reporting explains why glass breaks in these events, and it is not what most people assume. Wind pressure alone almost never breaks a sealed pane. An unlatched sash is caught by the wind and slammed repeatedly, and it is the impact that breaks the glass from the middle out. The click you hear when the cams seat is the whole safety mechanism.
| Load or action | Which direction it pushes | What resists it | Whether the lock point count helps |
|---|---|---|---|
| Positive wind pressure on an outward opener | Pushes the sash onto its stop, the same direction the lock pulls | The frame stop and the gasket | Barely. This is the easy direction |
| External negative pressure on an outward opener | Tries to lift the sash away from the frame | Every cam behind every keep, plus the hinges | Yes, and spacing is what decides whether any gap opens |
| The weight of the sash | Downward, always | The hinges, and the fixings behind them | No. Zero locking points carry the weight |
| Levering at the hinge side | Prying the sash off its hinges | Dog bolts engaging independently of the hinge pins | No. This is a separate component with a separate part number |
| The glass being attacked | Straight through the pane | The glass specification | No. No lock point count has ever stopped a pane |
A drive rod is straight. It lives in a straight groove, it drives cams along a straight line, and the keeps it drives into sit on a straight frame member. That single fact rules out a whole family of shapes, and it is worth knowing before the shape is drawn rather than after.
Thermal movement belongs in the same section. The Chinese window company's article requires three-dimensional adjustment on the keeps precisely because gaskets age and shrink and profiles expand and contract. A dark finish absorbs more than a pale one, and a large sash moves further than a small one, so a black swing door standing in full sun is a harder adjustment problem than the same item in a light colour on a shaded elevation. That is not a defect in the lock; it is the reason the adjustment has to exist and has to be used.
This is the boundary most marketing copy steps over. A UK guidance page describing PAS 24:2022 says the standard puts hinges, locks and handles through a timed forced-entry attack with screwdrivers, crowbars and hammers, and that to pass, the window must not open by more than 100 millimetres. Another UK knowledge base notes that BS 7950:1997 was the older enhanced-security standard for timber casement and sash windows, and that PAS 24:2022 is the current benchmark referenced in the Building Regulations for new dwellings.
Note what is tested. Not the lock. The window. The rating attaches to a specified assembly — frame, sash, hardware, glazing, and the way they are put together — and it is issued with a certificate number by an accredited laboratory. A count of locking points is an input to that assembly, not a substitute for it. The same knowledge base makes the point from the other side: hardware alone cannot bring a substandard frame up to the standard.
We would rather state this plainly than let a schedule be built on it. Of our three casement window series, none publishes a count of locking points. One names a transmission box and anti-pry lock points. One names its hardware brand and a cycle figure. One names an optional hardware upgrade and the hinge material. The only place a number appears anywhere in our published range is on a side-press sliding window, and it is printed without the sash size it belongs to.
That is not an oversight we are quietly hoping you will fill in from a competitor's catalogue. It is the correct behaviour, for the reason this whole article argues: a count printed without a sash size is not a specification, and we are not going to print one.
| Item | What we publish | What you still have to ask for | Why there is a gap |
|---|---|---|---|
| Lock point count on a casement | Nothing. No number is printed for any casement series | The count, the spacing, and the sash size both belong to | Because the count follows your sash, and your sash is not our number to invent |
| Locking geometry | A full-perimeter distribution statement on the inward-opening passive series | Whether the corners are driven, and whether the hinge stile carries points | A perimeter statement describes where, not how many or how far apart |
| Hardware brand | Named German and partner brands across the range | The specific mechanism reference, not the brand | A brand makes dozens of mechanisms with different counts and different ratings |
| Cycle figure | A cycle number for one series' hardware | What was cycled: the handle, the gearbox, or the whole sash under load | It is a rig figure, not a service life, and it says nothing about lock points |
| Engagement depth | Nothing | The lap, measured from a stated datum, and the adjustment range of the cam | We have not published it, so we cannot let you compare on it |
| Hinge-side security | Hinge material and construction on one series | Whether dog bolts are included, how many, and whether the keeps are on the order | Keeps are frequently a separate line item and can be omitted without anyone noticing |
| Operating force | Nothing | The handle torque or force at the largest sash in your order | It depends on sash size, glass weight, and how well the cams are set |
| Corrosion and shear figures | A salt spray duration and a shear value for one series | Confirmation of which series they belong to | They are coating and composite-bond results. They do not transfer to another series and they are not a lock rating |
| Security rating | Nothing | Ask your market's authority which standard applies, then ask who can be the applicant | Certificates are issued to a tested assembly and to an applicant in your jurisdiction. We can supply evidence, not the certificate |
Fill these in and send them with the drawing. The third column is what happens when the line is blank, and it is the only column most buyers read.
| Line | A useful answer looks like | If you leave it blank |
|---|---|---|
| Sash width and height, in millimetres, per opening | Actual manufacturing size, not the nominal call-out | Any count you are quoted is attached to somebody else's sash |
| Number of locking points, per side, at that size | Handle side and hinge side stated separately | You get a perimeter total that cannot be checked against any spacing rule |
| Largest gap between two adjacent points, in millimetres | One number, for the tallest sash in the order | You have the count but not the thing the count is derived from |
| Distance from the end point to the corner | A stated figure, in millimetres | The corners, where prying starts, have no stated protection |
| Mechanism type | Espagnolette with mushroom cams, or shootbolt, named | You cannot order a spare in year eight, because the two are not interchangeable |
| Mechanism reference, not just the brand | The maker's part code for the exact rod and gearbox | A brand name is not a part number and will not get you a replacement |
| Backset, rail length, rail width, cam height | Four numbers, in millimetres, plus inline or offset | The one dimension the industry standardised is the handle hole spacing, and it is not one of these |
| Engagement, measured from a named datum | A lap figure, plus the adjustment range of the compression cam | You cannot compare two suppliers, because they may be measuring different things |
| Keeps: material, fixing, and whether they are on the order | A named material, a fixing method, and a line on the packing list | The frame half of every locking point can arrive missing and nobody will notice until the handle is stiff |
| Hinge-side bolts: how many, and are the keeps included | A count per sash, and the keeps listed separately | The side an intruder actually attacks on an outward opener has nothing specified |
| Who performs the first compression adjustment, and when | A named party and a named stage in the programme | Every window arrives with its cams at the factory setting, and nobody ever touches them again |
| Security standard, version, certificate number, laboratory, configuration | Five fields, or a written statement that none applies in your market | You have bought a number of locking points and no rating at all |
| Ask this | This sounds like an answer and is not |
|---|---|
| How many locking points on my largest sash | "Eight points." Eight over what length, on which side |
| What is the largest gap between two points | "They are evenly distributed." Evenly over how many millimetres |
| Is it a multipoint lock | "Yes, of course." Every casement on the market has more than one point now |
| What hardware is it | "German hardware." Which mechanism, and what is its reference |
| How far does the cam engage | "It has mushroom cams." That is a shape, not a measurement |
| Are the keeps included | "Everything is included." Ask for the keep line on the packing list by name |
| What about the hinge side | "The hinges are heavy duty." Heavy duty does not engage into the frame |
| Is it tested | "The system is tested." Which standard, which version, which configuration |
| How hard is the handle to turn | "Very smooth." Smooth at what sash size, with what glass |
| Who adjusts the compression on site | "It is adjustable." Adjustable by whom, and is anyone booked to do it |
Not answerable as asked. Eight points spread over a 6 metre perimeter gives an average gap of 750 millimetres, which fails the spacing figure in the standard clause we quoted. Four points on a 3 metre perimeter gives 375 millimetres, which passes comfortably. The four-point lock is the better specified one. Once you know the length, the question answers itself, and you will find that a lot of published counts stop looking impressive.
They make it harder to lever the sash away from the frame at more places, which is real, and that is the limit of it. Security performance is tested on a complete assembly against a named standard and issued with a certificate number. A count is an input to that assembly. It is not a proxy for it, and a supplier who offers a count where you asked for a certificate has told you something.
Usually because nobody has worked it out for your sash, not because it is secret. The count comes from a hardware catalogue entry for a standard sash; the spacing comes from doing arithmetic on your actual one. If a supplier can produce the spacing after you send the sizes, that is a good sign. If they cannot produce it after two requests, they are quoting a catalogue, not engineering your order.
No. It usually means the geometry is wrong. A UK repair guide states that a correctly aligned lock should need no force to engage the handle, and that having to pull the sash inward to turn it means the keeps are too tight. Look for scrape marks on the keeps: that is where the cam is striking instead of seating. The usual repair is moving a keep by one or two millimetres, not replacing anything.
Not in any straightforward sense. The cams are fixed to a drive rod of a set length, the keeps are positioned to match them, and the rod length is bounded by the sash and by what is stocked. A longer rod, more keeps, and a re-adjustment of every cam on the sash is a hardware change, not a topping-up. A published source warns explicitly that surplus points raise operating resistance and that mis-adjustment puts points out of position.
More than the model name. Record the mechanism reference and the brand marking stamped on the faceplate, the backset in millimetres, the rail length, the rail width, the cam height, and whether the lock is inline or offset. A UK hardware specialist notes that multipoint locks are interchangeable between makes only when all of those match, and that swapping makes usually forces the keeps to be repositioned. Without that list, a failed gearbox in year eight becomes a new window.
