How the two differ in kind
A fence post is a cantilever pushed sideways. Wind arrives across the panel, travels through the rails into the post, and turns into bending concentrated right at ground level — and everything below that line exists to resist the rotation. Neither material is holding the fence up. The soil is. Both concrete and gravel do the same job of turning a slender post into a fatter buried body, so the sideways pressure is spread over more soil and the ground pushes back over a longer lever. The difference is what that fatter body is made of: one is rigid, bonded and watertight, the other is loose, unbonded and free-draining. Every argument on either side of this question is downstream of that single distinction.
The two camps are describing two different failures, and both failures are real. Concrete fails wet. Water runs down the post face, reaches the collar, and if the top of that collar is flat or dished, or the hole was filled with concrete right to the bottom so the water has nowhere to leave, it stands against the post at precisely the section carrying the highest bending stress. Timber rots there, steel corrodes there, and the capacity is gone before anything shows above ground. Gravel fails loose. A granular annulus resists only through particle interlock and the confinement of the soil pressing in around it, so it has to have been compacted hard, in shallow lifts, and it has to have stiff ground around it to be confined by. Miss either condition and the post has a little play at the top of its embedment — and play under a load that reverses with every gust patiently makes itself bigger.
Which of those stories applies to you is mostly settled before you pick up a bag or a shovel. Soil that drains hands gravel its entire argument; heavy clay hands it straight back, because a stone-filled hole in clay is not a drain, it is a sump with a post standing in the middle of it. A gate post, a corner and an end post carry a pull that never lets go, and granular fill creeps under a sustained one-directional load in a way that set concrete does not. A galvanised steel post does not rot the way timber does — it will eventually corrode at the ground line if water is allowed to sit there, but on a far longer clock — which takes most of the force out of the strongest case against concrete. So the honest answer is rarely one material for the whole boundary. It is concrete where the fence is asking for stiffness and gravel where it is asking for drainage, and a long run usually asks for both.
The factors that actually differ
| Posts set in concrete | Posts set in compacted gravel | |
|---|---|---|
| What actually resists the push | A rigid, bonded body. The collar and the post move as one piece, so the soil sees load the instant the fence does and the post has no play at the ground line — the stiffness is there on day one and does not need earning. | An interlocked granular mass. Load passes through stone-on-stone contact into the surrounding soil, which means the fill must be compacted hard in shallow lifts and confined by stiff ground before it resists anything at all. |
| Repeated wind load over years | Cyclic reversals do not loosen set concrete; whatever stiffness the collar had when it cured is the stiffness it still has, and any movement is the soil deflecting rather than the fixing slackening. | Every reversal works the stone a fraction. Well-graded angular material re-locks itself and holds; rounded stone or a lazy compaction job develops a wobble that each subsequent gust enlarges. |
| Water at the post face | The collar's profile decides the post's life. Crowned and sloped clear, it sheds; flat or dished, it holds a puddle against the most highly stressed section of the post for years. | Water runs straight through the fill and away — provided the ground around the hole will take it. The stone drains the hole; it cannot drain the soil. |
| What the surrounding soil has to be | Almost anything, with the caveat that soft ground limits how much the collar can achieve — the concrete spreads the pressure, but the soil still sets the ceiling. | Stiff enough to confine the stone and open enough to accept the water. Soft wet clay fails both tests at once, which is exactly where the gravel method is most often recommended and least appropriate. |
| Ground that freezes | The collar's shape decides whether frost gets a grip. Bell the top out and heave has a shoulder to pull against; taper it inwards, and take it deep enough, and there is nothing for the ice to hold. | Nothing is bonded to the post and the hole holds less water, so frost has less purchase — though a shallow embedment still lets the post creep upward season by season. |
| Getting it plumb, and keeping it plumb | One working window. Brace it, check both planes, and accept that when the mix goes off the position is final for the life of the fence. | Adjustable throughout, and still adjustable in five years — nudge, tamp, check, tamp again, and re-plumb a drifted post later by loosening the top of the fill rather than breaking out a footing. |
| When you can load it | The fence waits. Hanging a heavy gate or pulling wire tight against a collar that has not cured is how a post ends up leaning permanently on its first day. | Loadable as soon as the last lift is tamped, because there is no chemistry to wait for — the strength is entirely mechanical and entirely present. |
| The hole and the quantity you order | A narrower hole does the same work, and the buy quantity comes down again because the post is standing in part of it — a small deduction per hole that is worth having across a boundary. | Usually a wider hole for the same restraint, and the quantity moves the other way: you buy loose stone, lose some of it to compaction, and order it by the tonne rather than by volume. |
| Undoing it later | Every post becomes a buried boulder. Replacing one means excavating it whole or cutting the post off below grade and starting a new hole beside it — the most expensive thing on a fence to change your mind about. | Pull the post and the hole and the stone both stay. The replacement drops into the same excavation and gets tamped back in, which is why a boundary that might move is gravel's argument on its own. |
Which one, and when
Choose posts set in concrete when…
- Gate posts, corner posts and end posts — anywhere the load is sustained and one-directional. A gate's weight and a tensioned wire never let go, and granular fill creeps under a load that never stops.
- Tall solid fences with real sail area, where the wind load reverses relentlessly and any play at the ground line has decades to make itself worse.
- Clay and other slow-draining ground, where a stone-filled hole collects water instead of shedding it and gravel's headline advantage inverts.
- Galvanised steel or concrete posts, where the decay mechanism the whole case against concrete rests on is either absent or runs on a timescale long enough to stop deciding the question.
Choose posts set in compacted gravel when…
- Line posts on a permeable fence — post-and-rail, wire mesh, spaced pickets — where the wind passes through and the load per post is modest.
- Sandy, gravelly or otherwise free-draining native soil that will genuinely accept the water the hole hands it.
- Timber posts you want decades out of, since keeping standing water off the ground line is worth more to wood than to anything else in the fence.
- Anything you expect to move: an unsettled boundary, a temporary run, ground with services in it, or a fence you already know gets replaced before you do.
Now run your own numbers
This page holds no prices on purpose — a national average is wrong for almost every real project. Quantify both options with your dimensions and your local quotes.
Frequently asked questions
- Does concrete really rot fence posts?
- Concrete is not the agent — trapped water is, and concrete is merely very good at trapping it when it is shaped carelessly. Three details do the damage: a collar finished flat or dished so it ponds against the post, concrete taken right to the bottom of the hole so water arriving down the post face has no exit, and timber that was never rated for ground contact in the first place. Fix all three, by crowning the collar so it sheds clear, bedding the post on a compacted stone base so the hole drains from below, and buying the right treatment grade, and concrete-set posts last a very long time. It is also worth saying the other half plainly: gravel-set posts in clay stand in water too, and rot just as thoroughly. The material that surrounds the post matters far less than whether water can leave.
- How much gravel does a hole take compared with concrete?
- The two quantities are worked out in opposite directions, which is the practical reason this page hands you two calculators rather than one. The concrete figure shrinks: the post is standing in part of the hole, so what you buy is the hole less the post, and the post-hole calculator makes that deduction and takes your number of holes as an input, giving a total and a bag count in one go. The gravel figure grows: loose stone compacts down, so you need more than the finished hole volume, and the yard sells by weight, so the order ends up as a volume converted into tonnes at the bulk density of the grade you picked. One practical note on the second calculator — it is built for an area rather than a round hole, so give it the combined plan area of all your holes at your hole depth and let its compaction allowance do the rest.
- What kind of gravel — is pea gravel fine?
- Rounded pea gravel is the single most common mistake in this method. Round particles cannot interlock; they roll past each other under load and behave more like ball bearings than a structural fill, so the post is left with play from the day it goes in. What the method needs is angular crushed stone with a graded range of sizes, including fines, so the particles wedge against one another and the mass locks up when it is compacted. Compaction is the other half and it is not optional: shallow lifts, each one tamped hard before the next goes in, checking plumb as you climb. Dumping the stone in and prodding it with a bar produces a post that is loose immediately and looser every winter.
- Can I use both methods on one fence?
- Yes, and for most runs it is the right answer rather than a compromise. Concrete the posts that are being asked to be stiff — the gates, the corners, the ends, and any straight stretch long enough that the middle has no return bracing it — and set the line posts between them in compacted stone. Each material then does the job it is actually good at. The one thing worth thinking about before you mix them is the fence itself: rigid panels spanning between posts of noticeably different stiffness will concentrate movement at the joint between a stiff post and a softer one, so on a panel fence it pays to keep whole runs consistent and change method at a corner. Rail and wire fences absorb the difference without complaint.
- Does either method stop frost heave?
- Neither one exempts you, and depth is what actually decides it. In ground that freezes, the footing has to reach below the local frost line — a shallow post is lifted regardless of what surrounds it, and no amount of extra concrete or extra stone prevents that. Below that depth, the two behave a little differently. Frost grips the sides of a bonded concrete collar, so a collar that flares wider at the top hands it a shoulder to lift against, while one shaped narrower at the top resists being pulled out. Gravel puts less water in the hole and does not bond to the post, so it hands frost less to work with, but a short gravel embedment can still ratchet upward over successive seasons. Get the depth right first; the choice of fill is a second-order effect on this particular problem.