What Flooded Basement Cleanup Costs on Long Island

Search this and you will be given a range within about four seconds. Three to five thousand for a small job, ten to twenty for a bad one, more if it is finished. The ranges are broadly plausible and almost none of them says where it came from, which makes them impossible to argue with and impossible to rely on.

There is a different kind of answer available, and it is specific to this part of Long Island. FEMA publishes every flood insurance claim it has paid, down to the individual property, and the data can be filtered to the Town of Oyster Bay as its own community. That record does not tell you what a cleanup costs. It tells you something more useful, which is how wrong the idea of a typical figure is.

What the public record for this Town actually says

Filtered to the Town of Oyster Bay, FEMA's claims dataset holds 5,364 flood insurance claims for losses between 1978 and 2024. Of those, 4,658 resulted in a payment, and the payments come to $286,877,896 (FEMA, checked 2026-08-31).

The median of those paid claims is $33,176. The mean is $61,588.

Those two numbers describe the same set of payments and they are nearly twice apart, which is the first thing worth taking from any of this. When a mean sits that far above a median, the average is being dragged by a small number of very large values, and quoting it as what something usually costs is misleading in a specific direction: upward.

Then there is the reason the gap is so wide. Of the $286.9 million paid out here across almost fifty years, $250.0 million came from Hurricane Sandy. One storm in October 2012 accounts for 87.1 per cent of every flood insurance dollar ever paid in this Town, across 2,578 paid claims, with a median of $77,679 (FEMA, checked 2026-08-31).

Take Sandy out and the picture changes completely. For losses from 2013 onward there are 72 paid claims on the record, and their median is $7,137, with a mean of $16,630 (FEMA, checked 2026-08-31).

So the honest summary of half a century of insured flood losses in this Town is that an ordinary year produces a small number of claims with a median around seven thousand dollars, and that roughly once a generation something happens that is ten times worse and produces more than the previous thirty years put together.

That is not a price list. It is a much better description of the risk than any range, and it explains why the ranges disagree with each other so violently: they are averaging across two entirely different populations of event.

What this record does not measure, which matters as much

A number is only useful if you know what it counts, and this one has four real limits. None of them is hidden, and all four are the reason this post cannot end with a figure for your basement.

It is flood only. The NFIP pays flood claims, and a flood has a narrow published definition. A supply line that let go, a hot water tank that failed, an appliance that emptied itself into a finished basement, a drain that backed up: none of those is in this dataset at all, and between them they are the majority of wet basements. What you are looking at is the subset of water losses that arrived from outside the building.

It is what was paid, not what the work cost. A claim payment is settled after the deductible comes off and inside the policy limits. A loss larger than the coverage shows up here at the coverage amount. The relationship between these numbers and a contractor's invoice is real but loose.

It is insured properties only. A house with no flood policy that took four feet of water in 2012 appears nowhere in this record. That has a direction: the dataset understates how many properties flooded and says nothing at all about what the uninsured ones spent.

And it is nominal dollars spread across decades. A 1985 payment and a 2024 payment are added together as written. Over a period this long that flatters the older half of the record, which is one more reason the post-2013 figure is the more useful one for thinking about today.

There is a fifth thing that looks like a finding and is partly an artefact, which is worth saying out loud because it would be easy to make a meal of it. Of the 5,364 claims, 4,955 come from a single ZIP code on the south shore, which is 92.4 per cent of them (FEMA, checked 2026-08-31). That is a real concentration, and it says less about where water goes than it first appears. Flood insurance is required with a federally backed mortgage inside the mapped high-risk zones, and those zones here are overwhelmingly along the south shore bays. So the record shows where flood policies are held at least as strongly as it shows where basements get wet. The north shore's water problems are real and are mostly outside this dataset, which is its own kind of information.

Why the same job costs different amounts on the two sides of this Town

The Town runs from the Sound down to the bays, and the ground underneath those two halves behaves differently enough that it changes what the work is.

On the north side the land falls toward the water and the material under it is mixed. Water arrives at a building from uphill, during weather, and finds a way in at a particular place: a window well, a cellar entrance, a gap at the footing. The cleanup there is usually bounded, because the source stops when the rain does. The expensive part is often not the cleanup at all, it is the work that stops it happening again, and that work is outside rather than inside.

On the south side the ground is flat and the water table sits close under it. Water rises through the slab and the base of the walls rather than pouring in at one point, and the supply of it does not switch off when the sky clears. Two things follow for the bill. The drying takes longer, because the source has not actually been switched off, and equipment on hire is a daily cost. And the recurrence problem is a drainage project rather than a repair, which is the single widest variable anywhere in this subject.

That is why a range quoted for Long Island as a whole is doing something questionable. It is averaging a bounded job over a slope with an unbounded one over a water table, and reporting the middle as though it described either.

What actually moves the number

Setting aside the water, five things do most of the work, and they are the ones a contractor is trying to establish while they walk around.

What is down there. Finish is the whole game. A basement with a painted concrete floor, block walls and shelving is a drying job. A basement with carpet, an underlay, stud walls, insulation and a ceiling is a demolition job followed by a drying job followed by a rebuild, and the rebuild is usually the largest of the three. Floor area barely moves this. Materials move it enormously.

How long it was wet. This is the variable most under your control and the one people discover late. Wet material that dries quickly can often stay. Wet material that has been sitting comes out. How long before mold grows after water damage covers what the published window does and does not promise, and the short version is that speed changes what is salvageable rather than starting a clock.

What was in the water. Clean water from a supply line, ground water, and anything from a drain are three different jobs with different rules and different protective requirements. The category changes what can be cleaned and what has to be discarded, and it changes it regardless of how long ago it happened.

How long the equipment stays. Drying is billed by the day, with air movers and dehumidifiers sitting in the room, and the number of days is decided by the materials and by whether the source is genuinely closed off. This is where a quote and an invoice most often diverge, and where asking what happens if the readings are not down on schedule is a fair question rather than a suspicious one.

Whether one firm or three do it. Extraction, drying and reconstruction are sometimes one contract and sometimes three, and a quote covering one of them looks dramatically cheaper than a quote covering all three. A great many comparisons that seem to show a large price difference are actually comparing different scopes.

The contents question is worth a line of its own, because the record has something to say about it. Of everything paid out here, $251.2 million went on buildings and $35.7 million on contents, so contents are 12.4 per cent of the total (FEMA, checked 2026-08-31). Insurance and repair are not the same thing, but the proportion is a reasonable corrective to the instinct that the furniture is the loss. Mostly, the building is.

Where the national figures fit, and where they do not

It is worth knowing the scale of the thing, if only to see how little a national number tells you about a house.

FEMA reports that floods caused more than $8 billion of damage to homes and businesses nationwide in 2024, and that $3.8 billion of that occurred in communities not considered high risk (FEMA, checked 2026-08-31).

Neither number tells you anything about a basement. What they are good for is calibration: the second one is roughly forty-four per cent of the first, which means that in an ordinary year a large share of the national bill is picked up by households who had been told their area was not the problem. That is the population most likely to be reading a cost page with no flood policy in the drawer, and what a New York homeowners policy actually covers matters more to their final figure than any range on this page.

Checking any of this yourself

Everything above is derived from a public dataset, and one of the advantages of that is you do not have to take it on trust.

FEMA puts the same claims data behind a lookup where you can select a state and a county and see the dollars paid and the number of claims over any period you choose (FEMA, checked 2026-08-31). Nassau County as a whole holds 51,520 claims against this Town's 5,364, which is a reasonable sanity check on the scale of what you are reading here.

The Town-level figures in this post come from FEMA's API rather than that page, because the lookup works by county and this Town is a community inside one. The script that produces them lives in this site's repository, runs against the live API with no key and no account, and prints the dataset's own as-of date beside the results so that a figure which has moved is visible as having moved. The version in this post was run against the dataset as published on 3 August 2026.

Keeping your own number down

If your basement is wet right now, cost is the wrong first question, and the right one is whether it is safe to go down there. Federal guidance is to avoid wading in floodwater, which can be contaminated and can hide debris, and not to touch electrical equipment that is wet or while standing in water, turning the power off first only where that can be done safely (FEMA, checked 2026-08-31).

After that, the order that keeps the number down is unglamorous: stop the source, get the water out, and start drying the same day rather than the next one. What that involves is set out on flooded basement cleanup, and where the water came in because weather opened the building up, storm damage restoration is the closer description of the repair.

Then get more than one quote, and make them comparable before you compare them. Ask for extraction, drying and reconstruction as separate lines. Ask what is being removed rather than dried, and why the line was drawn where it was. Ask how many days the equipment is staying and what the plan is if the readings are not where they should be. A quote that cannot be broken into those parts is not cheaper, it is less specific.

And treat any figure you are given before somebody has looked at the room as what it is, which is a guess with a decimal point on it. This page has published every number it can stand behind, and none of them is the cost of your basement.

Where these figures come from, and how to check them

The claims figures are derived rather than quoted, so the derivation is published with them. FEMA also puts the same dataset behind a county lookup anyone can use, and the script in this site's repository re-runs the whole calculation against the live API.

• FEMA OpenFEMA dataset: NFIP Claims (v3 API) Checked 2026-08-31
• FEMA, National Flood Insurance Program Historical NFIP Claims Information and Trends Checked 2026-08-31
• FEMA, National Flood Insurance Program The Cost of Flooding Checked 2026-08-31
• FEMA Floods (Ready.gov) Checked 2026-08-31