Rust on the post: cosmetic, or structural?

The cleanest definitions we found are in the glossary of FHWA’s manual for bridge inspectors, which spends its life on exactly this question at a much larger scale:

That is the whole distinction, and the phrase “not yet” is the useful part of it. Surface corrosion is rust that has not taken material away. Once it has, paint does not put it back.

A third kind is worth knowing on a mailbox post specifically, and the same manual describes it in its chapter on steel rather than in the glossary: “Pack rust occurs between two mating surfaces; an increase in volume of rust over the original steel may create localized distortion and possibly cracking.” On a mailbox that means a joint — between a post and a bracket, or between a bolted plate and the post. A brown stain weeping out of a bolted joint, and a gap that was not there last year, is pack rust rather than a paint problem.

We are going to describe a physical test rather than give you a threshold, and we want to be clear about why. FHWA defines the concept, but every published framework for measuring it — condition states, corrosion grading scales — is written for inspected infrastructure, with calipers, ultrasonic thickness gauges and an engineer at the end of it. Nobody publishes a homeowner version. Any percentage or wall thickness we gave you would be our opinion wearing a source’s clothes.

What you can do without instruments:

• Look at the ground line first, not at the worst-looking patch. Soil, water and air meet there, and it is where a steel post is thinnest by the time anything shows higher up.
• Press hard with a thumb, then tap. Sound steel is unyielding and rings. Steel that has lost section gives, or sounds dead.
• Scrape a patch with something dull, a putty knife rather than a screwdriver. Rust that lifts off in flakes and leaves firm metal is surface corrosion. Rust that keeps coming, or a blade that goes through, is section loss.
• Push the post at the top. If the top moves and the ground does not, something in the post has gone. That is the same check as in the repaint or replace guide.
• Look at every joint and hole. A drilled hole, a cut end and a welded seam are where the coating was broken first, and they rust first for that reason.

Anything that gives, moves inside its own length, or lets a blade through is not a painting job, whatever the rest of it looks like.

A galvanized post does something a painted one does not: the zinc protects steel it is no longer covering. The American Galvanizers Association, the trade body for that coating, describes the mechanism this way:

How wide a bare stripe that covers, we are not going to tell you, because the same publisher gives two answers an order of magnitude apart on two of its own pages and reconciles neither. Take the mechanism and leave the number.

What a fully weathered-off coating means is a longer question than a dangerous one. FHWA’s manual for ancillary structures puts it plainly: “Total loss of galvanization may not be an immediate problem if there is no visible corrosion or section loss, but it may result in a shorter structure service life” (FHWA, Ancillary Structures Inspection Reference Manual (retrieved September 22, 2026)). Read that as: a dull grey post with no red on it is not an emergency, and it is the moment a coating starts earning its keep. The same section names the standard for putting zinc back, ASTM A780, Practice for Repair of Damaged and Uncoated Areas of Hot-Dip Galvanized Coatings, which is a different job from painting over it.

The standards that govern this are SSPC-SP 2, Hand Tool Cleaning, and SSPC-SP 3, Power Tool Cleaning — the two every industrial metal primer datasheet names as its minimum. Both define a prepared surface the same way, and the definition is narrower than most people assume: free of all loose mill scale, loose rust, loose paint and other loose material, and it is expressly not intended that adherent material be removed. The test for “adherent” in the standard is whether it can be lifted with a dull flexible putty knife.

So preparation is not about getting to bare metal. It is about removing everything the new coating would otherwise be stuck to rather than stuck on. Paint sprayed over loose rust has keyed into a layer that is already letting go.

Two manufacturers say the same thing on their own datasheets, which are free to read where the standards are not. Sherwin-Williams gives hand tool cleaning as the floor for its metal primer — “Iron & Steel- Minimum surface preparation is Hand Tool Clean per SSPC-SP2” — and adds the line worth keeping in mind before any repaint: “Recognize that any surface preparation short of total removal of the old coating may compromise the service length of the system.” (Sherwin-Williams, Kem Bond HS, PDS 125.21, 07/2021 (retrieved September 22, 2026)) Rust-Oleum’s rusty metal primer sheet says “STEEL: Scrape and wire brush or power tool clean to remove loose rust, scale and deteriorated previous coatings”, and twice warns not to use that product on galvanized steel (Rust-Oleum, Stops Rust Touch Up Rusty Metal Primer, SRT-22 (retrieved September 22, 2026)). Which matters here: a primer formulated for rusty steel is not a primer for a zinc coating.

The caveat the standard itself carries is the one worth passing on: heavily stratified rust or pack rust can adhere so tightly that a wire brush will not shift it, and even though it counts as tightly adherent by the putty-knife test, it makes a very poor surface to paint over. Tightly stuck and worth painting are not the same judgement.

A note on this section’s sourcing, because the rest of these guides link everything. SSPC-SP 2 and SP 3 are published by AMPP and are not free. We read the 2018 editions, and only on third-party hosts, because AMPP’s own site refused every request. The current editions are dated 2024 and we have not read them. We link only a publisher’s own page, so there is no link here. The section-2.2 definition quoted above was word-stable from 1982 to 2018, which is evidence and not proof.

It asks you to deal with it. USPS’s own Mailbox Improvement Week release lists “repainting a mailbox that may have rusted or has started to peel” among the tasks it asks homeowners to take on (USPS newsroom (retrieved September 22, 2026)), and its customer page recommends an annual check that includes tightening loose hinges and taking care of rusty or loose parts (usps.com, Mailbox Guidelines (retrieved September 22, 2026)). The reason it is your job rather than theirs is DMM 508.2.1.3: the purchase, installation, maintenance and replacement of mail receptacles “are not the responsibility of the USPS” (Domestic Mail Manual 508 (retrieved September 22, 2026)).

What we did not find, and looked for, is any USPS rule that mail stops because a box is rusty. The nearest thing is a notice: a carrier reports a non-conforming box and the postmaster sends a form asking the owner to fix it. That is in the USPS rules guide.

Then Virginia has rules about what goes back in, and they are short. The regulation, amended effective 19 November 2025, defines a breakaway structure as “a single four-inch by four-inch square or four-inch diameter wooden post or a standard strength, metal pipe post with no greater than a two-inch diameter” (24VAC30-151-560, as amended 2025-11-19 (retrieved September 22, 2026)). Build round wood to four inches: VDOT’s undated guidance sheet still says 4½, and where the sheet and the regulation disagree the regulation governs. The sheet is the source for the rest — embedded no more than 24 inches, and posts “shall not be set in concrete unless the design has been shown to meet VDOT breakaway standards” (VDOT, Mailbox Installation Guidelines (undated) (retrieved September 22, 2026)). The post materials guide has the rest of that sheet and the material trade-offs.