The ‘best’ way to brace a wooden door or gate

Got a sagging wooden door or gate?  Building a new one?  Not sure how you should brace it?  Confused about the conflicting advice you may have read elsewhere?  Read on.

Wooden doors and gates come in many different shapes, sizes and styles.  A very simple one would look something like this:


Unfortunately, if you build it that way, it will soon end up looking more like this:


The weight of the door/gate has transferred to a relatively small number of nails/screws that hold the wooden boards to the horizontal rails, crushed the adjacent wood fibres, and now the door/gate is sagging.  Your door/gate may already look like that.

To repair or prevent sagging requires that the door/gate be braced.  A brace is a diagonal piece of wood that takes a bit of the load off the nails/screws.

Tension Braces

Many doors/gates have tension braces.  They look like this:


A tension brace works by transferring load from the outside end of the bottom rail to the inside (hinge) end of the top rail.  It ‘pulls’ the weight of the gate up to the top hinge.

Compression Braces

The other — and now far more common — type of brace is the compression brace.  They look like this:


A compression brace works by transferring the load from the outside end of the top rail to the inside (hinge) end of the bottom rail.  The weight of the gate ‘rests’ more on the bottom hinge.  Load is transferred something like this:


So, which is better?  How should you brace your door/gate?

Tension vs Compression

Let’s kick off by just clarifying one thing:  Any type of bracing is better than no bracing at all.  Pretty-much all doors/gates require some form of bracing, but depending on size, shape and weight, you can sometimes get away with very little bracing and the type of bracing doesn’t really matter.  Narrow and/or light doors/gates fall into the ‘it probably doesn’t matter’ category.

If your door is wide and/or heavy, however, then you are better off with a compression brace.  A compression brace is easy for most DIYers to construct to a satisfactory standard, and is hard to screw up.  Just remember:

  • Compression braces are suitable for doors where the angle between the bottom rail and the brace is greater than 45°
  • Make sure both ends of the brace have full contact with the rails
  • Put (ideally two) nails/screws through each board into the brace

Note:  It is entirely possible to tension brace a wide and heavy gate — if you have a decent amount of experience with different types of joints and you know what you are doing.  The average person, however, doesn’t have that experience, and doesn’t have the required carpentry skills…

The biggest advantage that compression braces have over tension braces is the way that loads are transferred.  In compression, loads are distributed over the end cross-section of of the brace as well as the screws/nails — that greatly reduces the overall rate of fibre compression (which means it will sag less and last longer).  Because the loads are transferred to the bottom hinge, and the bottom hinge is closer to the ground, the gate ends up more stable — it bounces up and down less.  Gate posts supporting tension-braced gates also tend to bend/twist over time — because the load is transferred to the top hinge (higher from the ground, where it exerts more torque).

Why The Contradictions?

If compression braces have such a clear advantage, why do some sources still recommend tension braces?

Mainly due to historical reasons, but also because in a very limited number of scenarios tension is still better.

In the good old days, farmers and homesteaders would cut their own timber/lumber with axes and saws.  The wood would be ‘green’.  Green lumber (i.e. freshly cut or anything that hasn’t been kiln-dried or seasoned for a couple of years) contains a large amount of water. As it dries it shrinks. The shrinkage will occur in all dimensions, but most noticeably along its length. The lumber will get shorter. If you are using green lumber, a tension brace is ideal because the shrinking wood will help pull the bottom rail up.

Since nearly all lumber used nowadays is purchased from a store, and has been seasoned/kiln-dried, it contains very little water. If you live in a climate with a wet season, the lumber will absorb water and expand. It will get longer. If you build using dry lumber, a compression brace is ideal because the expanding wood will push the top rail up.

In the good old days, steel wasn’t used for gate posts.  Timber posts were sunk into bare soil.  Since most gates were outside and exposed to the elements, rain would hit the gate from both sides.  A compression brace would channel water that hits the ‘back’ of the gate down towards the gate post — where it would saturate the soil, encourage rot, and result in the premature failure of the post.  A tension brace would channel water away from the gate post — helping to keep the soil there dry and prolonging the life of the post.

Modern gate builders often use steel gate posts sunk into concrete, and don’t expect their gates to last a generation, so don’t care where the water goes when it runs off the gate.  It’s not even a consideration.  If the brace is out of the weather (e.g. the inside of a barn, shed or house door) then it’s also not an issue.


When would a tension brace be acceptable, or even advisable?

  • Narrow doors/gates
  • Light doors/gates
  • When using green lumber, which includes:
    • lumber you have sawn/milled yourself
    • (store-bought or reclaimed) lumber that has been exposed to rain
  • In exposed situations where the gate post is subject to rot
  • If you intend to always keep the door/gate painted/oiled (the brace won’t lengthen during the wet season because water won’t get absorbed in any appreciable amount)
  • If you want — or are prepared — to use a turnbuckle brace, then tension is the best/only way to go

In all other situations, the average DIYer would be better off installing a compression brace.  If you’re not sure, install a compression brace.

Anyway, I hope you found some of that interesting or useful.  Cheerio!

27 thoughts on “The ‘best’ way to brace a wooden door or gate

  1. I see a lot of modern looking horizontal cladded gates with out a brace. I’m planning on building one, but Curious if solid joinery will compensate for lack of bracing, or how these gates are designed to eliminate the need for a brace.

  2. Many ‘modern’ product are designed with planned obsolescence in mind — they are designed to fail and thereby promote replacement and new sales. Products are not inherently ‘better’ by virtue of being ‘modern’. That said, good design, materials, manufacturing and assembly can — definitely — deliver a door/gate that doesn’t sag appreciably over time and thus does not need a dedicated bracing element. Expect to pay a pretty penny for one, though.

    Compression and tension braces, as described above, are easy and cheap ways to keep low and medium-quality doors and gates — installed by the average Joe — functioning the way they are supposed to. Turnbuckle tension braces are also a relatively easy and discreet way to fix a gate that promised more than it delivered.

  3. After trawling through the Internet trying to research which direction would be best to add a brace for a 3’7″ gate ( slightly wider than normal I suppose), this is the ONLY detailed explanation I have found to help me decide.

    I will now be adding a compression brace .

    Thank you for this information.

  4. A person built my gate with only a turnbuckle. Its already sagging. I will use the turnbuckle. But do I need a compression brace also?

  5. Bobbie, turnbuckle-braced gates and fences usually sag a little in the first year as wood fibres compress and components settle in. The good thing about turnbuckles, though, is that they can be adjusted. If you haven’t already, just tighten the turnbuckle until the gate is back in the desired position, and then leave it for another year. It is not uncommon to have to turn the buckle quite a few times in the first year, a few more times in the second year, a small amount in the third year, and then very little (if any) after that.

    If you have reached the limit of how far the turnbuckle can be tightened, you can always stick something under the free end of the gate to support it, loosen the turnbuckle as far as you can, then disconnect one of the cables, shorten it, reconnect it, then tighten up the turnbuckle again.

    Without seeing the gate, it is impossible to know if it has a fundamental design flaw that is causing it to sag. If a “professional” installed the gate, however, then it’s probably not fundamentally flawed, and just needs some tightening. I don’t advise adding any more bracing until you’ve had the gate (and kept adjusting the turnbuckle) for a minimum of three years. If it still keeps sagging significantly after that, the reason it is sagging should have become obvious (e.g. wood tearing, plate slipping, post bending). Then you can either fix the problem or install more bracing.

  6. I have more of a question than a comment.
    I have a 10′ x 12′ shed in my back yard. It has a single door that measures 52″ x 72″. It has 3 horizontal braces on the outside. It’s pretty much a rectangle with one horizontal 2×4 in the middle. There are 3 large “T” hinges holding the door to the shed. The door has sagged about 3/4″. With the door being divided into two sections, what would be the best type of bracing to help prevent the door from sagging again ? Once I get the door square again, of course.
    Thanks for your assistance.

  7. Karl, your best bet is to install one compression brace in each section (from outer top corner down to inner bottom corner). Something like this:

  8. Would have a compression as well as a tension brace, ie a cross brace with the intersection rebated and securely screwed together further strengthen the gate? Thanks

  9. Wilson, if you are suggesting a cross-brace for Karl’s situation, then I don’t think that would be a good idea at all. He describes his door as being a rectangle with a horizontal 2×4 through the middle. If you tried to add a cross-brace to that then you would have three pieces of wood all crossing at the same point and they would need to be cut very thinly indeed (very deep rebates). Since the strength of a brace is proportional to cross-sectional area, any reduction in thickness (i.e. any form of rebating) reduces the strength of a brace.

    If you weren’t talking about Karl’s situation, and are talking about cross-braces in general (no horizontal framing in the middle), then the same logic applies, but you end up with 50% depth rebates, and lose half of your bracing strength, instead of two-thirds. Not quite as bad, but still bad.

    Although I (personally) like the look of cross-braces, the reality is that they don’t (mechanically) work as well, which is why I don’t use them.

    A tension brace is good if your lumber is green/wet, so that when it dries it pulls tighter. If you make both parts of a cross brace out of the same lumber, or planks from the same stack of lumber, then they will have similar moisture levels. If the moisture level is ‘high’ then your compression brace will shrink and weaken the door. If the moisture level is ‘low’ then the tension brace will expand and weaken the door. So, no matter what the moisture level, some part of the bracing system is going to malfunction. A lose-lose situation.

    Because I don’t want any part of my doors or gates undermining the rest, I like to keep their designs simple. Stick with one type of bracing if at all possible. If the visual appearance of double compression bracing on a door is a problem, I’d rather add a thin veneer of some sort to the sheltered inside. Something like a 3mm sheet of plywood wouldn’t add much weight, but would make it look nicer (IMHO) — and might even help stiffen the door up a bit.

    Note: All of the above refers to wooden cross-braces. If you have metal bracing then a welded cross-brace is perfectly fine. Metal is not affected by moisture levels in the same way that wood is.

  10. What Ahmed said.

    I (a physics major in college) had convinced myself that t didn’t matter which way the brace ran, but Maine Cabin Masters said to have it run to the bottom hinge. So I searched the internet for an explanation, found several not-very-helpful or not-very-convincing how-to’s, and then found this. Excellent explanation. So my *next* z-back door will have a compression brace, unlike my first one.Thank you!

  11. I’m building a small wooden gate which will be wider than it is tall. Should I use a compression brace, but keep the angle 45 degrees such that the brace will not span the entire width?

  12. Zach, if you are going to make the effort to brace a gate then you may as well brace the full width. In the case of gates that are twice as wide as they are high, the best approach is to switch from one gate to two. For gates that are not quite that wide, or when two gates aren’t a viable option, the simplest approach is to use two compression braces side-by-side instead.

    The first brace extends from the bottom rail near the bottom hinge up and out diagonally until it meets the top rail. Drop straight down from that to the bottom rail and start your second brace there. The second brace then goes up and out until it meets the top rail near the outside (free) edge of the gate. Basically, a saw-tooth pattern — and essentially the same pattern that gives trusses their great strength.

    Assuming your gate is symmetrical, the top of the first brace and the bottom of the second brace should contact their respective rails near the mid-point of your gate. That should make both brace angles identical — something most folk prefer the look of.

    Gates that are three times as wide as they are high should really have steel bracing, or a turnbuckle tension brace. Gates four (or more) times as wide as they are high should be made entirely of steel, not wood, with saw-tooth bracing all the way — basically they become trusses at that point. You can make quite wide gates entirely out of wood, but only if you design and build them like commercial and industrial trusses. That’s a non-trivial task beyond the capability of most DIY folks.

  13. Thanks a lot Time. After some other research I was leaning towards the 2 parallel braces as well.

  14. Hi Tim,

    Thanks for this useful article. I love that you are still answering questions on it after 2 years. I am planning to make a hidden gate out of featheredge fencing so that it is indistinguishable from the rest of the fence on the public side similar to the example at

    The hinges and bolts will be fixed to a small section of rail cantilevered from posts 8″ either side of the gate with a 1:4 cantilever backspan. I was planning to use a double-z compression brace to reinforce the gate as there are three horizontal rails and no vertical framing. However, the person who installs these hidden gates suggested I run the diagonals from the top and bottom of the hinged side of the gate, meeting at the middle on the bolt side of the gate. After searching online it seems this is called a ‘British brace’ although I have also seen it with the hinges and bolts on the opposite side to what is described above. I wondered if you could explain what are the advantages of this system, if there is any particular benefit to me for my specific usage and what side the hinges and bolts should go on?

    Kind Regards,


  15. Greg, there are precisely zero technical advantages to the “British Brace”. Indeed, assuming that all the braces start with the same moisture level, regardless of whether the weather is wet or dry, one brace will always always be undermining¹ the integrity of the gate. A “Double-Z” (double-compression) brace, made from seasoned/dried timber, does not do this — so is technically superior.

    The reason the British Brace exists is nationalism. The shape of a British Brace forms one half of the Union Jack (the flag of the UK). If you have two gates, with the same design on both, you get a full Union Jack. That’s it. That’s why the design exists. No other reason.

    As has been the case throughout all of human history, when design is driven by nationalism instead of functionality, you end up with an inferior product. The British Brace is no exception.

    The farmers of old recognised the liability, which is why back in the day you would usually only see the design on barn doors — sliding doors — that were hung from a top rail, not hinged from the side. Sliding doors barely need bracing at all, so a cosmetic brace design isn’t a liability in such a use-case.

    Physics recommends that a side-hinged gate with three horizontal rails should be supported by two compression braces. Go Double-Z.

    ¹ Assuming you start with dry timber, in the wet season the timber will expand and lengthen. The top brace of a British Brace — extending from the inner side of the top rail to the outer side of the middle rail — will try to push the gate down towards the ground. This is the exact opposite of what you want bracing to actually do. All bracing should work to counter the effects of gravity, and should either pull or push the outer side of the gate upwards. The rule-of thumb for side-hinged gates: ‘Up and in’ is good; ‘Down and out’ is bad.

  16. Thanks Tim,

    I really appreciate it. I wish there were more forums like this as helpful as yours. Now to try and persuade them to do it the proper way, I know ‘professionals’ don’t like being told what to do!



  17. Brilliantly useful article. Thank you.
    May I ask your opinion please. When using a Z brace (compression) is it functionality important that the angle of both braces is the same? The situation is where the middle horizontal rail is not centered i.e. it’s nearer the top or bottom rail. The angle of both braces will be kept greater than 45 degrees so that rule won’t be broken. Cosmetically it’s not important to me as the back of the door will be skinned with 3mm ply.
    Thanks in advance

  18. Dave,

    Each section (top and bottom) of a Double-Z brace design is, effectively, independent.

    Load transferred by the top brace should be directed through the middle hinge into the post. Load transferred by the bottom brace should be directed through the bottom hinge into the post. This holds true provided that the bottom/inner end of each brace rests on the part of a rail that is also directly attached to a hinge — which is almost always the case.

    Brace angle symmetry is not required when sections are independent.

    PS: A Double-Z design with three horizontal rails but only two hinges (i.e. the middle rail is not hinged) is sub-optimal design that should be corrected. Hopefully this does not apply to you, and all three of your rails are hinged. If only two hinges support three rails, then the top and bottom sections are not truly independent and… well… the physics becomes surprisingly complicated at that point.

  19. Thanks Tim,
    Yes sorry that’s exactly what I meant, a Double-Z design with three horizontal rails 🙂 You make a good point about the need to hinge the middle rail!

    I didn’t mention it but I’ll be using stiles. I assume the addition of stiles (so becoming a framed, ledged and braced door) would lessen the need for three hinges because the load from the top brace would transfer down the stile to the bottom hinge? Or could the same load, unsupported by a middle hinge bow the stile over time and cause its own problems!?


  20. Hello Tim, I hope this note finds you in good health.
    I have a gate that is 36″ by 36″. That is, the rails are 36″ long and the slats are 36″ long.
    The difference between the bottom of the top rail and the top of the bottom rail is18 3\4 “. When I cut my brace on a mock up using old lumber for the rails and the brace in order to make my mistakes there instead of when I use the pressure treated lumber, I noticed that the angle from the bottom hinge to the top latch was quite a bit less than 45 degrees, which would make the transfer of power quite weak.
    I decided to make two braces and attach them equidistant from each other. In this way, the angle was greater than 45 degrees. I haven’t put it together yet because I wanted your opinion on whether it was a good idea to use two braces instead of one long brace.

    Thanks Tim

    James in Ottawa, Canada.

  21. James, apologies for the delayed response — was out of town and offline for the good part of a week.

    If what you are planning is something like the following image, then you’ll be fine:

    Note that the top of the inner brace and the bottom of the outer brace are connected to the same vertical board/slat. While this is optimal from a load transfer point of view, some folks prefer them to align with adjacent boards — usually because they have an even number of boards, not an odd number. Adjacent boards are acceptable if this is the case.

  22. David, stiles (vertical timbers on the inside and outside edge of a door or gate) do not have the positive effect that most folks think they do. If the stiles are in addition to the normal boards — which is typical, and appears to be the case by your use of the word ‘framed’ — then your gate will sag slightly more than a gate with no stiles at all. The reason is simple: Weight. You are adding mass to the outer edge (the worst possible place as mass there generates the most torque). The inner stile accomplishes nothing, as it is directly connected to all of your hinges. Load can’t be preferentially transferred down to lower hinges because, to do so, it would have to bend the hinges higher up. Since metal beats wood, that doesn’t happen — load just distributes over all hinges evenly.

    From a structural point of view, constructing a frame that looks like the number 8 on a calculator’s segmented display, and then cladding one side with planks will stiffen the gate. This makes it more resistant to things like wind and mechanical forces like animals trying to slip out, or trespassers trying to force the gate open. If you want added security, it’s not a bad idea. It makes a gate more secure at the expense of added weight.

    To offset the additional weight, folks sometimes add small braces to the outer corners. Do an image search for ‘gate corner brace’ to see what I mean.

    At this point complexity has gone up a few notches, and we’ve strayed from the original intent of this post: Guidance for DIYers with average carpentry skills on how to make simple and functional gates that won’t sag.

    ‘Bowing’ is not something you should be concerned about. That (unfortunately) happens when folks don’t pay attention to the grain of their boards, and too many boards end up with the same, non-parallel side grain. If your boards have parallel side grain your gate won’t bow. If your boards have non-parallel side grain, and you alternate their direction, then warping forces will cancel out and your gate won’t bow. ‘Randomly’ selected and assembled boards (which is what typically happens when folks aren’t even aware of the bowing issue) rarely end up in a configuration that bows the gate.

  23. Thanks Tim: Excellent explanation.

    I never thought about that i.e. “Note that the top of the inner brace and the bottom of the outer brace are connected to the same vertical board/slat. I did, however, upon reviewing your responses, see the explanation for Zach.

    BTW, How to add an image? Copying and pasting don’t work.

    James in Ottawa, Canada

  24. James, WordPress comments are limited to text. You can’t paste images into a comment. What you can do, however, is upload an image to your own website, or some image-hosting site, and then include the URL for that image in your comment. Put the URL on a line by itself and it will render without a hitch.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s