Most homeowners have never looked underneath their own deck. Ten minutes with a flashlight and a screwdriver will tell you more than any sales visit — here's the same sequence we run on every structural inspection, in order of what actually fails. Print it, walk your deck, and you'll know whether you're looking at nothing, a repair, or a real problem before anyone quotes you a dollar.
The ledger (where the deck meets the house)
What to check: From underneath, find the horizontal board bolted against the house. Look for metal flashing over its top edge, real bolts or structural screws (not just nails), and dry, solid wood.
When to worry: No visible flashing, rust streaks, soft or dark wet wood, or a ledger attached with nails alone. Industry post-mortems trace the large majority of catastrophic deck collapses to this one connection — anything suspicious here is a call-a-pro finding, not a watch-it finding.
Footings and posts
What to check: Posts should sit on concrete footings via metal brackets — not buried in dirt, not resting on patio blocks. Sight down each post for lean; look at the footing tops for cracking or heaving.
When to worry: Posts in ground contact (they rot from the bottom, invisibly), a footing that's lifted or tilted with frost, or daylight between post and bracket. Frost heave is Colorado's signature deck disease.
Joists and framing
What to check: Underneath, probe a few joists with a screwdriver — especially near the ledger and any stains. Solid wood resists; rot gives. Check that joist hangers have nails in every hole and no orange rust bloom.
When to worry: A screwdriver that sinks in easily anywhere, hangers missing fasteners, or joists pulling away from the ledger. Blackened wood that's still hard is usually cosmetic; soft wood never is.
Railings and posts
What to check: Grab the top rail and push hard, at several points — code expects a guardrail to resist a 200-pound load in any direction. It should feel like part of the structure, not furniture. Check that balusters are tight and spacing is under 4 inches.
When to worry: Any visible movement at the post base. A railing that wobbles today fails on the day someone actually falls against it — and post connections are the most common failure we find on older and DIY decks.
Stairs
What to check: Walk them heavily. Check the stringers (the sawtooth boards) where they meet the deck and the ground, the tread fasteners, and the handrail's steadiness. Eyeball the geometry too — risers over 7¾ inches or treads under 10 inches fail code.
When to worry: Bounce or flex mid-flight, stringers resting on bare dirt, treads cupped enough to hold water, or a handrail you wouldn't trust with your full weight. Stairs are where most deck injuries actually happen.
The surface
What to check: Scan for cupped, split, or spongy boards, popped fasteners, and gaps that have closed up (drainage matters). Note whether wear is cosmetic — graying, checking — or structural.
When to worry: Soft spots underfoot mean the problem is below the boards, not in them. Widespread splitting on a 20+ year old wood deck usually means the surface is done even if the frame is sound — which is exactly what re-decking is for.
Scoring it: repair, re-deck, or replace
Rough triage from what you found — the same logic behind every inspection we run:
The money logic: a re-deck over sound framing costs meaningfully less than a full replacement — as of 2026, about $20–35 per sq ft to re-deck over sound framing versus $40–70 per sq ft (plus $1,500–3,500 of tear-out) to rebuild in composite. An honest structural verdict is worth real dollars, which is why ours comes as a written report, free.
Not sure what you're looking at?
A free on-site structural inspection gets you the repair / re-deck / replace verdict in writing — no judgment, no pressure.
When to run this
Once a year, ideally in spring — Colorado's freeze-thaw cycle does its damage over winter, and you want to find what moved before the deck's busiest season. Run it too when you're buying a house (an unpermitted or failing deck is a classic inspection-report surprise), and before any big party puts thirty people on a structure built for ten. The odds aren't trivial: of the roughly 40 million residential decks in the U.S., industry groups estimate about half are past their intended service life.