Here is a number that surprises people. When Portland’s city auditor examined its own permit reviews, the finding was that more than 80 percent of standard building permit plans required corrections, and most of those went through several rounds.
Eighty percent. These are not homeowner sketches. These are sets prepared by professionals, and they still come back.
So either most plans are full of mistakes, or something else is going on. It is something else.
The plans are not wrong. They disagree with each other.
Ask people who prepare plans for a living what actually triggers corrections, and the same answer keeps coming up. It is rarely a missing sheet, and it is rarely a bad design. It is two documents in the same set that no longer describe the same building.
The window schedule lists one glass rating. The energy report was run with another. Both documents exist. Both were done carefully. They just describe two different buildings, so the set comes back.
Or the electrical diagram shows a 200 amp breaker while the panel schedule says 225. Or the framing plan never picked up the engineer’s revised foundation. One firm that reviewed hundreds of correction letters found this pattern again and again: not absence, not error, but disagreement.
How does that happen to careful professionals? Simple. Designs change while sets are being drawn. Somebody moves a wall, upgrades a window, revises a beam, and the change lands on three sheets but not the fourth. Each sheet is right about the building it describes. They just are not describing the same one anymore.
The reviewer is actually several reviewers
The other thing homeowners rarely see: your plans are not read by one person checking one code. In Portland, a residential set can pass through zoning, life safety, structural, site development, environmental services, forestry, water, and transportation as separate reviews, and every one of them has to sign off before the permit moves.
That is why a set can be structurally flawless and still sit, held up by a stormwater question or a tree. And it multiplies the disagreement problem: each reviewer reads a different slice of the set, so a mismatch between slices is exactly what gets caught.
It gets stranger: the fix itself can be rejected
Portland publishes a list of what it calls common rejection items on correction resubmittals, and almost none of them are about the building. Revised areas not clouded on the drawings. Sheets missing a date stamp. Comments not answered one by one. The wrong pages uploaded.
Read that again: you can fix the actual problem and still lose weeks because of how the fix was packaged. And each round has a real calendar cost. Portland’s target for re-review alone is 5 to 7 days, which sits on top of however long the revisions took to prepare.
There is one more trap in the response round. Fix the setback by moving a wall, and the wall move changes a dimension on two other sheets. Answer the comment, create a new disagreement. Practitioners say the projects that spiral are almost never the ones with the hardest problems. They are the ones where each fix was made in isolation.
What actually improves your odds
Knowing all this, the picture of a clean submittal changes. It is less about brilliance and more about agreement. A few things consistently separate sets that pass from sets that cycle:
Settle the site facts before anything gets drawn. A current survey, the real setbacks, the easements, where the utilities actually run. Practitioners who review correction letters say a large share of trouble started before drafting did, with a site assumption that turned out to be wrong.
Don’t submit half-finished. Sending the architectural set in while the structural calculations are still catching up is a known way to buy an extra round.
Freeze the design, then check the whole set against itself. Does the energy report match the windows actually drawn? Does every sheet the wall move touched get the move? The single most valuable pass is the boring one: reading the set the way the reviewers will, as separate slices that have to tell one story.
Answer corrections as a set, not as a list. Every comment answered where the reviewer can find it, every change clouded, and then the whole-set check run again, because the fix that creates a new conflict is the classic second-round story.
The honest expectation
Nobody prevents every correction. Cities read the same drawings differently, and practitioners working across many jurisdictions will tell you an identical set can sail through one town and cycle in the next. A correction round is a normal part of how permits work.
What is not inevitable is the avoidable round: the site fact nobody verified, the energy report nobody reconciled, the response package that fixed the problem and failed the paperwork. The difference between one round and four usually lives there, in whether the pieces of the set agree with each other.
That agreement does not happen by accident. Someone has to be checking for it the whole way through.
Sources: City of Portland Auditor, Building Permit Review Audit (2021). Portland Permitting & Development, review timelines and correction submittal requirements. City of Phoenix PDD, Residential Building Plans Prelog Administrative Review. Practitioner reporting compiled from residential design and permit consulting sources, 2024 to 2026. Last checked August 2026.