How Deep Should a Fence Post Be?

No building code tells you how deep to set a fence post

This is worth establishing properly, because almost every page on this topic implies otherwise. We read and can name the sections: IRC R401.1, R401.4, R401.4.1 and Table R401.4.1, R403.1, R403.1.4 and R403.1.4.1; IBC 1806.1, 1806.2 and Table 1806.2, 1806.3.1, 1806.3.3, 1806.3.4, 1807.1, 1807.2, 1807.3 and its subsections; and the International Swimming Pool and Spa Code 305.2.1. The word fence appears in the model building codes essentially only in the permit exemption list.

The reason is scope. IRC R401.1 states that "The provisions of this chapter shall control the design and construction of the foundation and foundation spaces for buildings." A fence is not a building. And the permit exemption cuts the fence out of the system entirely: the IRC exempts "Fences not over 7 feet (2134 mm) high" from permits, language confirmed in the adopted codes of Massachusetts, Michigan and Clear Creek County, Colorado.

Two consequences follow, and both matter more than the rule of thumb.

• Your city may still have its own number. Permit-exempt in the model code does not mean unregulated locally. Cook County, Illinois requires a permit "for fences constructed of any material at any height" and mandates that "Post holes must be dug to a depth of 42-inches." Lombard, Illinois sets a 36 inch minimum and requires a post-hole inspection before you pour. Woodland, California requires a permit over 6 feet and states "Fence posts must be set in concrete." Locally mandated depths we found run from 24 to 42 inches. Call your building department before you dig, because this is the only number that can actually fail an inspection.
• A pool fence is a different question. California amends its carve-out to read "Fences, other than swimming pool barriers, not over 7 feet high," so a pool barrier needs a permit there even under 7 feet. But note what the pool code itself regulates: the International Swimming Pool and Spa Code sets barrier height at 48 inches and ground clearance at 2 or 4 inches depending on construction, and specifies no post depth, diameter or footing at all.

The one-third rule, and why four versions of it disagree

"One third of the post height" is the most repeated number in fencing. It is not a code requirement, not a consensus standard, and not an engineering result. It is a bagged-concrete manufacturer instruction, and the manufacturers do not state it the same way twice.

Read the four phrasings carefully. One is a third of the exposed height, one is a third of the overall post, one adds six inches on top, and one gives a range up to a half. On a single 6 foot fence built on a standard 8 foot post those produce 24, 30, 32 and up to 42 inches. The arithmetic there is ours; the wordings are theirs. Same fence, same post, an 18 inch spread depending on which bag you happened to buy.

Nobody publishes a derivation or a test basis for any version of it. We looked at Quikrete, Sakrete, ASTM, USDA NRCS, the Chain Link Fence Manufacturers Institute and several agricultural extension services. The rule has no stated engineering origin anywhere we could find.

One confusion to head off, because readers who go digging will hit it. IBC 1807.3.2.1 contains its own one-third, defining the allowable lateral soil-bearing pressure as "based on a depth of one-third the depth of embedment." That is a term inside a force equation, not the origin of the rule of thumb. They are unrelated, and the resemblance is a coincidence.

The one published depth standard in fencing

There is exactly one, and it is written for chain link rather than wood: ASTM F567, the chain-link fence installation standard. Its footing rule is reproduced verbatim in the Chain Link Fence Manufacturers Institute product manual and in Merchants Metals' specification section 32 31 13, in identical words:

"Minimum footing depth, 24 in. (609.6 mm) plus an additional 3 in. (76.2 mm) for each 1 ft. (305 mm) increase in the fence height over 4 ft. (1220 mm)."

That is a straight line rather than a ratio, which is the interesting part. It starts at a fixed 24 inches and adds a fixed 3 inches per foot of fence, so it does not scale the way the one-third rule does.

The depths in the table are our arithmetic on the published rule, not a table printed in the standard. And one caution we want to be precise about: neither reproduction states a maximum depth. You will see a 60 inch cap attributed to F567 online; we checked both documents specifically for it and both are silent. Do not rely on a cap we could not find.

One boundary worth naming before the detail. A deck post is not a fence post and the code treats them completely differently: a deck footing is sized off an IRC table by tributary area and soil bearing, because it carries load, and the deck footing size calculator reads that table. Everything on this page is about a non-structural post that only has to resist wind and stay upright. If your post holds up a roof, a pergola or a deck, stop reading here and size a footing.

Two honest limits on using this for a wood fence. It governs chain link, where the post is a steel pipe rather than a 4x4. And the standard is paywalled, so what we have is two independent manufacturer reproductions rather than the standard itself. Even so, 30 inches for a 6 foot fence is the most defensible single number available, and it lands in the middle of the manufacturer spread above.

Frost depth, which overrides the rule of thumb

In a cold climate this is the number that matters, and it is not a national figure. IRC Table R301.2 is printed blank in the model code. Its footnote instructs that "The jurisdiction shall fill in the frost line depth column with the minimum depth of footing below finish grade," so the frost line that applies to you is a local lookup, not something you can derive.

The spread is enormous. The University of Arizona publishes 0 inches below 4,000 feet of elevation and 24 inches above it. Seattle publishes 12 inches. Columbus, Ohio 32 inches. Blue River, Colorado 40 inches. Minnesota Rules 1303.1600 sets "five feet in Zone I and 3-1/2 feet in Zone II," which is 60 inches at the deep end. A rule of thumb that gives 30 inches is adequate in Seattle and half of what Minnesota wants.

What actually lifts a post, and why it is not what you think

The common explanation is that ice forms under the post and pushes it up. The measured mechanism is different and it acts on the sides. The classic research is from the National Research Council of Canada, Penner and Burn, August 1970, which describes it as: "The upward thrust of the growing ice lens is transmitted to the foundation unit to which the frozen soil is bonded by adfreezing." The frozen soil grips the concrete collar and lifts the whole assembly.

The forces are not small. That work measured roughly 6,000 pounds of uplift on a 3.5 inch steel post in clay at 3.5 feet of frost penetration, about 12.5 psi of adfreeze shear, with 10 psi measured on a concrete block wall and 18 psi given as a design guideline. Two caveats we will state rather than bury: that research is Canadian and it concerns foundation columns, not fences.

Which raises a question nobody authoritative has answered. Read the code footnote again: it defines the frost line as a minimum footing depth. A fence post in a concrete collar is not a bearing footing, and the adfreeze mechanism above acts on the side of a slender element rather than under it. We could not find any authority addressing whether frost-depth rules apply to a fence post the way they apply to a footing. The two facts point the same direction, but joining them is our inference and not a sourced conclusion. The practical advice everyone gives, get below the frost line, is probably right. It is just not established.

Worth knowing that trade sources get this badly wrong in both directions. Sakrete's own project guide, as published, tells installers to dig "about 6' below the frost line." That is almost certainly a typo for six inches, and it is a good illustration of why a bag instruction is not a specification.

How wide: the rule, and the code minimum nobody cites

"Three times the post width" is the rule of thumb, and unlike the depth rule it is reasonably well corroborated. Quikrete publishes "three times the width of the post." USDA NRCS Maryland specifies "a hole that is at least 12 inches deep, with a diameter that is at least three times the diameter of the post. (For example, a 4-inch diameter post shall have a minimum 12-inch diameter hole filled and set with concrete.)" NRCS Texas requires a minimum 12 inch diameter where concrete is used. Three independent sources, same rule.

Two published figures sharpen it, and neither appears on any other fence page we have seen.

The IBC gives an absolute minimum

IBC 1807.3.3, governing backfill around an embedded column, requires that "The hole shall be not less than 4 inches (102 mm) larger than the diameter of the column at its bottom or 4 inches (102 mm) larger than the diagonal dimension of a square or rectangular column." A 4x4 is 3.5 inches actual, so its diagonal is 4.95 inches, which puts the code minimum at 8.95 inches, call it 9. A 6x6 at 5.5 inches actual has a 7.78 inch diagonal, so 11.78 inches, call it 12. Those are our computations off the published rule and the dressed sizes in the American Softwood Lumber Standard PS 20, which gives nominal 4 inch as 3.5 inches dry and nominal 6 inch as 5.5 inches dry.

The fencing standard wants wider than you think

ASTM F567's diameter rule, again from the CLFMI and Merchants Metals reproductions in identical words, is "Minimum footing diameter four times the largest cross section of the post up to 4.00" (101.6mm) O.D. and three times the largest cross section of post greater than 4.00" (101.6mm) O.D."

Note it is four times below a 4 inch post, not three. Applied to a 3.5 inch 4x4 that gives 14 inches, which is wider than the 10 to 12 inches the whole trade digs. That is our arithmetic applied across a material boundary, so treat it as a direction rather than a requirement: the one real standard in fencing is more conservative on diameter than the rule of thumb.

Depth beats width, and here is why

If you are choosing where to spend effort, dig deeper rather than wider. The reason is in the code equation. IBC 1807.3.2.1 gives embedment for the unconstrained case as d = 0.5A{1 + [1 + (4.36h/A)]^1/2} with A = 2.34P/(S1 b), where b is the "Diameter of round post or footing or diagonal dimension of square post or footing" and S1 is the allowable lateral soil-bearing pressure.

Because A is inversely proportional to b, required depth scales roughly as the inverse square root of diameter once the hole is reasonably deep. Our derivation, shown so you can check it: for 4.36h/A much greater than 1, d is approximately 1.044 times the square root of A times h, and since A is proportional to 1/b, d is proportional to b to the power of minus one half. Doubling the hole diameter cuts the required depth by only about 29 percent. Concrete volume, meanwhile, goes up with the square of the radius, so a 12 inch hole holds more than twice what an 8 inch hole holds. Deep and narrow is cheaper than wide and shallow for the same resistance, which the bag counts make obvious.

Soil, and the table that actually applies

Most fence pages cite IRC Table R401.4.1 for soil. That is a mistake and we want to correct it plainly. Table R401.4.1 is titled "Presumptive load-bearing values of foundation materials" and contains vertical bearing pressure only. It has no lateral column. A fence post does not fail by punching down into the ground, it fails by rotating through it, so vertical bearing is the wrong property.

The table that does apply is IBC Table 1806.2, which publishes a lateral bearing pressure in pounds per square foot per foot of depth:

Clay is the worst case on the table by a wide margin, at a twelfth of bedrock and half of sandy gravel. That is the published basis for "posts in clay need to go deeper," and it is a much better reason than the folklore version. Two provisions make the table usable. 1806.3.3 permits the pressure to be "increased by the tabular value for each additional foot of depth to a value that is not greater than 15 times the tabular value," which is the formal statement that depth compounds. And 1806.3.4 permits "lateral bearing pressures equal to two times the tabular values" for isolated poles "not adversely affected by a 1/2-inch motion at the ground surface due to short-term lateral loads." A fence that moves half an inch in a gust is a fence, not a failure, so that doubling is the provision closest to what you are building.

Where the standard rules stop working, the published advice is to switch method rather than to adjust the numbers. USDA NRCS Maryland: "In extremely wet or very sandy soils, and in cases where posts cannot be set to the specified depth, the posts of permanently installed fences must be set in concrete to secure them." Note the direction of that, because it is the reverse of what you usually read online: USDA treats wet soil as a reason to use concrete, not a reason to avoid it. For shallow bedrock, NRCS Texas allows that "In rocky soils a 5 1/2 foot post can be used and set 18 inches in the ground," against its own 24 inch standard. On drilling into ledge, epoxy anchors or surface-mounting to rock we found nothing authoritative at all, only forum advice.

Two related calculators if the ground is the problem. Where water is collecting against a fence line rather than draining, the slope calculator works out the fall you need over a run. And if you are putting a gravel layer in each hole, the gravel calculator gives the tonnage for the whole fence rather than making you guess at a bag count.

Wind, and why a solid fence is a different problem

Depth and diameter both come out of a lateral force in the code equation, and on a fence that force is wind. The thing that decides how much wind your fence catches is not its height, it is how solid it is.

The Chain Link Fence Manufacturers Institute publishes a wind load guide that makes this fence-specific. The guide "is based on the assumption of a solid panel of fencing and uses multiplication factors for various percentages of free area," and for 9 gauge, 1.75 inch mesh the factor is 6.4. In other words chain link sheds roughly six times the load that the same area of solid privacy fence has to carry. If you are reading a depth figure off a chain link source and building a board fence, you are reading a number sized for a fraction of your load.

For gate and corner posts, everyone says deeper and wider. Two USDA NRCS state specifications actually publish it. California 2019 sets line posts at 24 inches and brace and gate posts at 3 feet, with a hole "at least 6 inches larger in diameter than the diameter of the post" and a 12 inch minimum where concrete is used. Vermont sets line posts at 30 inches and corner and gate posts at 36. Chain link sizes gate posts up by leaf width, and CLFMI notes that "Pipe terminal posts are generally one size larger in outside diameter than the line posts."

We could not find a published gate-post depth or diameter for a residential wood fence. The agricultural specifications above are the nearest published figures, and the pattern across all of them is consistent: the gate post goes about 20 to 50 percent deeper than the line post. Treat that as the shape of the answer rather than a standard.

For a sanity check on absolute depth from outside fencing entirely, two agencies publish numbers for sign posts, which are a similar slender element in a similar load case. FHWA describes wood sign posts as "buried about 30 to 36 inches deep," and Missouri DOT's engineering policy guide requires "a minimum of 36 inches into the ground." Both state an absolute depth rather than a ratio, which is itself a comment on the one-third rule.

Setting the post: the steps, and the two details sources fight about

The mechanics are not complicated. Dig, add the drainage layer if you are using one, set and brace the post plumb, place the concrete, check plumb again, leave it alone.

1. Dig to your depth and diameter. A clamshell digger for a few holes, a powered auger for a run. Keep the sides as vertical as you can; a hole that flares at the top wastes concrete and does nothing for resistance.
2. Compact the bottom. This is the step with real engineering behind it and almost nobody does it. ANSI/ASAE EP486.2, the shallow post and pier foundation standard, requires that "All disturbed soil at the base of a hole must be compacted to a magnitude consistent with the soil bearing capacity assumed in design." Loose spoil at the bottom of the hole is how a post settles.
3. Set the post and brace it before any concrete goes in. Plumb on two adjacent faces, then brace with scrap and stakes. Fast-setting mix gives you minutes, not hours.
4. Place the concrete. For the dry-pour method that fast-setting mixes are designed around, the powder goes in dry and the water goes on top. Our Quikrete fence post guide covers that method step by step and the bag counts that go with it, and dry pour concrete covers where else the method holds up, which is a shorter list than the internet suggests.
5. Check plumb once more in the first minutes, while there is still a hair of movement, then stop touching it.
6. Wait before you load it. USDA NRCS Maryland requires that "Fence wire shall not be attached to posts until at least 5 days after setting the posts in concrete," which is far more conservative than the 4 hours a fast-setting bag allows. The bag figure is about the concrete; the USDA figure is about the whole assembly taking tension.

Does gravel in the bottom do anything?

Six inches of gravel in the bottom of the hole is near-universal advice. Here is what is actually published. The instruction traces to a concrete manufacturer rather than to any research body: Quikrete's own step 2 is "Pour 6" of gravel or crushed stone into the bottom of the hole. Compact and level the gravel using a post or 2 x 4." Its stated purpose on the same page is rot protection, and Quikrete frames it as one of the "popular methods" rather than as a requirement. Quikrete's own product data sheet for the same mix does not mention gravel at all.

The commonly quoted "4 to 6 inches" has no primary source we could find. Six inches, from Quikrete, is the only traceable figure.

There is no published evidence that the gravel layer extends post life. There is published doubt, and it comes from the right place. A 1979 extension publication written by four USDA Forest Products Laboratory scientists, De Groot, Feist, Eslyn and Gjovik, "Protecting wood fences for yard and garden" (University of Wisconsin-Extension A3052), makes the point that crushed rock improves drainage away from the post in light soils but can trap a pool of water around the base of the post in clay. We are paraphrasing rather than quoting, because the only host for that document blocks automated reading and we will not put quotation marks around a sentence we have not read ourselves. The position is attributable and the logic is sound: drainage only works if the water has somewhere to go. A gravel pocket in clay is a bathtub.

Also worth knowing that no USDA NRCS fence specification we checked, Maryland, Texas, North Dakota or Nebraska, mentions gravel or crushed stone at all. And one extension source argues the other way outright, on frost-jacking grounds, warning that rocks put back in the hole can act as levers to push the post out of the ground. In free-draining soil the gravel is cheap insurance. In clay it may be doing the opposite of what you intend, and nobody has measured it either way.

Above grade or below? Four authoritative sources, four answers

This is the detail that decides whether water sits against your post, and it is genuinely unsettled. Every source below is primary and they do not agree.

Those cannot all be followed. Note in particular that USDA NRCS is instructing installers to bring concrete above grade and slope it, around posts that the same document requires to be treated for ground contact. That is a current federal specification taking the opposite position from the 1979 extension advice that gets quoted everywhere, and the extension advice is the one that offers a reason while citing no data for it.

Our read: both camps are trying to stop water sitting against the post, and they disagree about whether a concrete collar sheds it or holds it. Nobody has measured it. If you keep the concrete below grade, cap it with soil that slopes away. If you bring it above grade, slope it hard and make sure it is bonded to nothing that will crack away from the post. The failure to avoid is the flat collar that ponds, which is what you get by doing neither deliberately.

Does concrete rot a wood post? What the research actually says

This is stated as fact on most pages about fence posts. It deserves a careful answer, because the honest one is more interesting: the claim as usually put is not supported, not cleanly contradicted, and in the form that matters to a homeowner, unstudied.

First, where posts actually rot

Three independent sources agree, and the answer reframes the whole argument. The University of Kentucky's wood preservation publication locates it precisely: "Utility poles are most prone to decay in the groundline zone (from 6" above grade to 18" below grade) and the pole top." Oregon State's 1960 post farm report, which tracked real posts to failure, says the same in different words: "The most vulnerable section of a post extends from a short distance above to some distance below the ground surface. This zone usually has a sustained favorable supply of moisture and air necessary for existence of these destructive agents."

Moisture and air. That is the mechanism, and it is why the deep end of a buried post is not the part that fails. The University of Kentucky publication lists what decay fungi require, and the fourth item is the one that matters here: "Adequate oxygen - fungi cannot live in water-saturated wood." Decay also needs a wood moisture content "of about 30% (the generally accepted fiber saturation point of wood)" and a temperature "between 50°F and 90°F."

Follow that through and the folklore has a problem. A mechanism that genuinely held the post saturated would slow decay, not accelerate it. A mechanism that keeps it intermittently damp and aerated would accelerate it. Nobody has published measurements showing which of those a concrete collar actually produces.

The one peer-reviewed study, and what it really tested

There is a single relevant paper: Nicholas, Rowlen and Milsted, "Effect of Concrete on the pH and Susceptibility of Treated Pine to Decay by Brown-Rot Fungi," Forests 2020, 11(1), 41. It opens by noting the gap: "Treated wood timbers employed in ground contact are often installed with a cement collar to firmly fix the structural wood post in place. Few prior studies have determined the effect of concrete on decay efficacy on treated wood, however."

Cement contact raised the pH of the outer few millimeters of wood sharply, from the mildly acidic 5 to 6 range to 7.3 to 9.3. The decay results then split by preservative. Untreated wood showed no consistent effect either way. Amine copper azole showed no meaningful difference. Amine copper quat improved in decay resistance after cement exposure. Micronized copper azole got worse. The paper's own conclusion is that "decay susceptibility may be unaffected, reduced, or enhanced, depending on the particular copper/organic wood preservative employed."

State the scope limit with it, because it is the whole point. That study tested the chemical effect of cement contact on treated pine. It did not measure moisture retention and it did not compare the service life of posts in concrete against posts in soil. It settles a narrower question than the one everybody is arguing about.

What nobody has done

We went looking for a field trial comparing post service life in concrete, in soil and in gravel. There is none. No service-life comparison, no moisture-content measurements inside a concrete collar. Several extension publications assert the mechanism; none of them cites data for it. That absence is the real state of this question, and it is worth more to you than a confident answer would be.

What does demonstrably matter: the treatment

The variable with real published evidence behind it is not the backfill, it is what the post is treated with and to what retention. The AWPA Use Category System is the published framework. A sawn fence post in the ground is UC4A, ground contact, general use; a picket or rail above the ground is UC3B, above ground, exposed. The Western Wood Preservers Institute publishes that mapping and USDA NRCS Maryland specifies exactly those two categories for exactly those two parts. Note that the IBC requires a higher grade, UC4B, for embedded structural posts and poles, so if your post carries a roof rather than a fence panel, that is the code answer.

The service-life numbers are the argument. Four independent datasets, spanning 1960 to 2013, agree on the magnitude. Untreated pine posts last about 3 years in the South and 3 to 6 in the Midwest and Northeast. Untreated durable species do far better: Oregon State's post farm recorded western juniper at 27 years, black locust 24, redwood 23, western red cedar 22 to 23, against Douglas-fir at 4 to 7. Pressure-treated pine is reported at 30 to 40 years anywhere in the United States, and a Forest Products Laboratory severe-site study estimated 40 to 60-plus years for several preservatives.

So the honest hierarchy is: treatment first, drainage detail second, backfill material a distant third. A properly treated UC4A post is going to outlast the fence it holds up whatever you set it in. An untreated one will rot at the groundline in a few years whatever you set it in.

Concrete, gravel or foam

Three methods get argued about. The published evidence is lopsided, and not in the direction the internet suggests.

What the code permits as backfill

IBC 1807.3.3 lists exactly three options for the annular space around an embedded column: concrete "with a specified compressive strength of not less than 2,000 psi" in a hole at least 4 inches oversize, "clean sand" compacted "by tamping in layers not more than 8 inches in depth," or controlled low-strength material. Gravel is not on that list. Foam is not on that list.

Be precise about scope: that section governs embedded columns in structures, and a residential fence under 7 feet is outside the permit system entirely. It is the engineering benchmark, not a rule you are legally bound by. But it is the only published statement of what a professional is allowed to backfill a post with, and two of the three internet favorites are absent from it.

What expanding foam actually claims

Polyurethane post-setting foam is sold as a concrete replacement. We read both the US and Canadian data sheets for Sika PostFix, the market-leading product, and the gap between what is claimed and what is published is the story.

For every performance property that decides whether a post stays upright, the data sheet publishes a claim and no number and no test method. There is no ASTM or ISO reference anywhere in either sheet, and the two sheets disagree with each other on service temperature and mixing time.

The limitations are more telling than the omissions. The sheets state that the material is "not intended for setting deck posts" and, on the US sheet, "Do not use for structural posts or posts exposed to a lot of sideways load." Sideways load is the entire job of a tall privacy fence post.

If you are weighing which bag rather than which method, types of Quikrete and Sakrete mix types compare the products, and what psi concrete you need covers where strength stops buying anything.

Compare that with the bag of concrete it is sold against. Quikrete's Fast-Setting data sheet publishes compressive strength to ASTM C39: 1,000 psi at 24 hours, 2,500 psi at 7 days, 4,000 psi at 28 days. Numbers, and a standard they were measured by. That contrast is the honest core of this comparison.

On head-to-head performance: we found no independent testing comparing foam, concrete and gravel for setting posts. Everything available is trade content and video. If someone shows you a comparison, check who paid for it.

Setting in compacted gravel

Gravel is not in the code list, but there is published engineering on tamped granular backfill, from utility pole practice. USDA Rural Utilities Service Bulletin 1724E-200 states that "If an aggregate backfill is used, an ultimate skin friction between 250 and 1000 psf may be possible," against the warning that "If the soil is wet or subject to frequent wettings, an ultimate skin friction not greater than 100 psf should probably be assumed." It also specifies the technique: backfill "placed and compacted in shallow layers (approximately 6 inch layers)," with each layer compacted "until the tamp makes a solid sound as the earth is struck," and notes that "Power tamping of the soil is preferred."

Those are skin friction figures for direct-embedded utility poles, not fence post pullout, and we are not going to let them drift into a claim about fences. What they do establish is that compacted granular backfill has a published engineering basis and wet native soil is much weaker. The USDA NRCS specifications tamp native soil in 4 to 6 inch lifts (Texas) or 6 inch lifts (Maryland) and never mention gravel.

The practical summary. Concrete is the only option with published strength data and code recognition, and it widens the effective post, which is most of why it works. Compacted gravel has real engineering behind it in a related application and is easier to pull out when a post fails. Foam is fast and the manufacturer excludes it from exactly the load case a privacy fence imposes. Nobody has tested them against each other.

How far apart, and how many posts

Spacing decides your post count, which decides your total concrete, so it belongs here. No building code specifies fence post spacing, for the same permit-exemption reason as depth. What governs it is stated outright in one federal sentence. USDA NRCS Maryland, on wooden board fence: "Posts shall be spaced a maximum of 8 feet on center to accommodate rail lengths of a maximum of 16 feet."

Spacing is set by the material, not by the engineering. Rails come in 16 foot lengths, so posts go at 8 feet. That is the whole derivation, and it is published.

Published maximums vary enormously by fence type, from the same NRCS specifications: conventional woven wire 10 feet, barbed wire 16 feet, high tensile smooth wire 16 feet, or 30 feet with battens, and electric high tensile up to 60 feet on level ground. For chain link, Michigan State's construction standard requires line posts "equally not exceeding 10-foot on center."

The 6 foot spacing you will see recommended for tall privacy fences has no primary source we could find. It appears only in trade content, where it is offered as the conservative choice. The 8 foot figure is real, published, and justified. If you want closer spacing on a tall solid fence, the wind argument above is the reason, but no standard gives you a number for it.

Count posts for a straight run as the fence length divided by the spacing, plus one for the final post, plus an extra at every corner and both sides of every gate. The fence post concrete calculator takes the post count and gives you the whole fence in bags, and its worked example runs a 96 foot fence with a corner and a gate end to end.

Sources

Code text is quoted from jurisdictions that publish their adopted codes, since the model codes themselves are paywalled. Where we could not read a document ourselves we say so and paraphrase rather than quote.

Codes and standards

• 2021 Seattle Building Code chapter 18, reproducing IBC Table 1806.2, sections 1806.3.3 and 1806.3.4, and 1807.3.2.1 and 1807.3.3
• 2018 Seattle Residential Code chapter 4, reproducing IRC R401.1, R401.4, R401.4.1 and Table R401.4.1
• Grant County, Washington handout of IRC R105.2, work exempt from permit
• Chain Link Fence Manufacturers Institute product manual, section 32 31 13 and Merchants Metals specification 32 31 13, the two independent reproductions of the ASTM F567 footing rule
• CLFMI Wind Load Guide, on solidity and free-area factors
• Michigan State University construction standard 323113, chain link fences and gates
• ANSI/ASAE EP486.2, Shallow Post and Pier Foundation Design, on compacting the base of the hole

USDA and agency specifications

• USDA NRCS Maryland, MD382 fence specifications, 2016, the source of the spacing sentence, the sloped concrete requirement, the wet soil rule and the 5 day wire rule
• USDA NRCS Texas, Fence Code 382, 2015, on crowned concrete, tamped lifts and rocky soils
• USDA Rural Utilities Service Bulletin 1724E-200 section 12, foundation stability of direct-embedded poles
• Western Wood Preservers Institute, specifying with AWPA Use Categories and the AWPA U1 Use Category excerpt

Research on decay and service life

• Nicholas, Rowlen and Milsted, Effect of Concrete on the pH and Susceptibility of Treated Pine to Decay by Brown-Rot Fungi, Forests 2020, 11(1), 41, the only peer-reviewed study on concrete and post decay
• University of Kentucky, Wood Preservation, 2016, the groundline zone and the four requirements for decay
• Graham and Miller, Service Life of Treated and Untreated Fence Posts, Oregon Forest Research Center, December 1960, the post farm service life data
• De Groot, Feist, Eslyn and Gjovik, Protecting wood fences for yard and garden, University of Wisconsin-Extension A3052, the source of the below-grade concrete advice and the clay soil warning, paraphrased here because the only available host blocks automated reading
• Lebow, Woodward, Kirker and Lebow, Long-term durability of pressure-treated wood in a severe test site, 2013
• Penner and Burn, Adfreezing and frost heaving of foundations, National Research Council of Canada, CBD-128, August 1970

Product data

• Quikrete, Setting Posts in Concrete and the Fast-Setting Concrete Mix data sheet
• Sika PostFix US product data sheet and the Canadian data sheet