Landslide Risk in the US: How FEMA Scores It and How to Prepare

Written by Dvir · Data: FEMA National Risk Index v1.20 · Updated August 4, 2026

Landslides are the National Risk Index’s most underestimated hazard — they occur in all 50 states, cause an estimated $1–2 billion in annual damage nationally, and kill 25–50 Americans in a typical year, yet they carry no dedicated insurance market and often no warning system. The 2014 Oso, Washington slide killed 43 people in a valley mapped as hazardous decades earlier; the 2018 Montecito debris flows killed 23 in the first big rain after a wildfire. In the NRI, landslides (hazard code LNDS) account for roughly $250 million in national expected annual loss, with measurable risk in all 3,143 counties.

What counts as landslide risk

The NRI’s landslide hazard covers the downhill movement of rock, soil, and debris — from slow-creeping slumps that crack foundations over years to debris flows, the fast, deadly variety: slurries of mud, boulders, and trees moving at 30+ mph down channels. The recipe is almost always the same three ingredients: steep slopes, weak or saturated material, and a trigger — usually intense or prolonged rain, sometimes earthquakes, snowmelt, or human earthwork (undercut slopes, leaking pipes, added fill).

The most predictable subtype is the post-wildfire debris flow: burned soil sheds water like pavement, so for 2–5 years after a fire, even a modest downpour on a burn scar can launch a flow. This is now a routine, forecastable emergency in the West — Montecito was exactly this.

How the NRI measures it

FEMA combines each county’s exposure in landslide-susceptible terrain (buildings and population), annualized landslide frequency, and historic loss ratios into an Expected Annual Loss scored 0–100 against all counties. Note that the underlying loss records are widely considered incomplete — landslide damage is often filed under the triggering storm — so the NRI’s totals here are conservative.

Where the risk concentrates

The top of the ranking is the Pacific coast: Santa Barbara County, CA leads (score 100, ~$51M expected annual loss — Montecito’s county, logging ~60 slide events/year), with Snohomish County, WA (Oso’s county, ~$25M), San Bernardino County, CA, Monterey County, CA (Big Sur’s perpetually sliding coastline, ~130 events/year in the data), and Island County, WA (Puget Sound’s collapsing coastal bluffs). Beyond the top five, high ratings follow the Appalachians — eastern Kentucky, West Virginia, western North Carolina (where Helene’s 2024 rains triggered thousands of slides) — plus Colorado’s canyons and Alaska.

How to read your county’s score

Landslide risk is the most parcel-specific hazard in the index after wildfire: county scores flag regions, but your risk lives in your slope, your soil, and what’s uphill of you. The questions that matter: Are you on, above, or below a steep slope or in a canyon mouth? Has the area burned recently? Do your state or county publish landslide-susceptibility maps (Washington, Oregon, California, and North Carolina do)? And know the insurance reality: standard homeowners policies exclude earth movement — coverage requires a separate “difference in conditions” policy, and few homeowners have one.

What actually reduces the risk

  • Learn the warning signs on your property: new cracks in foundations, driveways, or soil; doors and windows sticking; tilting trees, poles, or fences; new springs or seeps on a slope; bulging ground at a slope’s base. These are how slow failures announce themselves — get a geotechnical engineer promptly.
  • Control the water. Most residential slides are water-triggered: keep gutters, downspouts, and drains discharging away from slopes, fix leaking irrigation and pipes, and never let runoff concentrate onto a hillside.
  • Respect burn scars. If you live below one, treat every NWS debris-flow warning as an evacuation order for the 2–5 years after the fire — flows arrive too fast to outrun.
  • During intense rain in steep terrain, stay alert to the sounds of a flow (cracking, rumbling, boulders knocking) and move up-slope or to upper floors; canyon bottoms and channel mouths are the kill zones.
  • Before buying in hill country: check susceptibility maps, look for repaired cracks and retaining walls on the property, and ask about slide history in disclosures — mitigation after purchase (retaining structures, dewatering) costs tens to hundreds of thousands.

The slides that changed what we map

Oso, Washington, March 22, 2014. A hillside above the Stillaguamish River collapsed without warning and ran across the valley floor, burying a neighborhood and killing 43 people — the deadliest single landslide in US history. The site had failed repeatedly before, had been studied, and appeared in geological reports going back decades. Houses had been permitted below it anyway. Oso is the reference case for the central problem of this hazard: the information frequently exists, and it does not reach the people making the decisions. Washington’s subsequent lidar mapping program, which has since revealed tens of thousands of previously unrecorded landslide deposits statewide, is a direct consequence.

Montecito, California, January 9, 2018. Weeks after the Thomas Fire burned the slopes above the town, a brief but intense rain cell dropped roughly half an inch in five minutes on the burn scar. Debris flows carrying boulders the size of cars ran through the community, killed 23 people, and destroyed or damaged hundreds of homes. The warning had been issued; the evacuation zones did not match where the flows actually went. Montecito is why post-fire debris flow is now treated as a distinct, separately warned emergency in the West.

La Conchita, California (1995 and 2005). The same coastal bluff failed twice, the second time killing ten people in a community that had already been buried once. It is the clearest available illustration that landslide sites repeat, and that living below a known failure is a decision, not an accident.

Western North Carolina, September 2024. Hurricane Helene’s rainfall on steep, saturated Appalachian terrain triggered thousands of landslides and debris flows, destroying roads, bridges, and homes across a region that most people — including many residents — did not think of as landslide country. It expanded the national mental map of this hazard well beyond the Pacific coast.

Thistle, Utah (1983) was economically instructive rather than deadly: a slow-moving slide dammed a river, destroyed a rail line and a highway, and inundated a town. It became one of the costliest landslides in US history and a standard case study in how a slow failure can be as expensive as a fast one.

Triggers, timing, and where the risk lives

Landslides need three things: steep or oversteepened slopes, weak or saturated material, and a trigger. The trigger is what gives this hazard its seasons.

  • Winter rain, West Coast (December–March). Atmospheric rivers saturating slopes over days. The damaging events usually come late in a wet sequence, when antecedent soil moisture is already high — the third storm, not the first.
  • Post-fire debris flows, first rains after a burn (typically October–March in the West). Burned soil develops a water-repellent layer and loses the roots and litter that held it, so a burn scar sheds water almost like pavement. Rainfall intensities that were harmless before the fire — on the order of half an inch in fifteen minutes — can launch a flow. The elevated risk persists roughly two to five years until vegetation re-establishes.
  • Tropical and convective rainfall, Appalachians and interior East (summer and autumn). Hurricane remnants and stalled thunderstorms on steep, thin-soiled slopes.
  • Spring snowmelt, Rockies and Intermountain West. Slow saturation from below, producing slumps and earthflows rather than fast debris flows.
  • Earthquakes, anywhere with seismic hazard. Shaking triggers slides across enormous areas at once; in the 1964 Alaska earthquake, ground failure caused much of the property damage.
  • Coastal bluff erosion, Puget Sound, the California coast, and the Great Lakes. Wave undercutting plus groundwater, producing failures that consume yards and eventually houses.
  • Human triggers, year-round. Cutting into a slope toe, adding fill or a swimming pool at the crest, leaking irrigation lines and pipes, removing vegetation, and routing roof runoff onto a hillside cause a large share of residential slides. This is the category you control.

The National Risk Index’s landslide ratings should be read as conservative. Landslide losses are chronically under-recorded, usually filed under the storm that triggered them, so the underlying loss history is incomplete almost everywhere.

What actually reduces the risk on a property

Water management is the whole game for most residential slides.

  • Get roof and surface water off the slope. Gutters and downspouts discharged into a solid pipe that carries water past the slope, not onto it. Never let a downspout, driveway swale, or sump discharge concentrate onto a hillside crest.
  • Fix leaks immediately. Leaking irrigation lines, pool plumbing, water services, and septic drainfields are among the most common triggers of suburban slope failures, precisely because they saturate ground slowly and invisibly.
  • Do not oversteepen or overload. Cutting into the toe of a slope for a patio, or adding fill, a pool, or a heavy structure at the top, changes the stability calculation. Both need engineering, not intuition.
  • Keep deep-rooted vegetation. Roots contribute real mechanical reinforcement in the upper soil. Clear-cutting a slope for a view is a documented way to destabilize it, typically two to five years later as the roots decay.
  • Engineered solutions when warranted. Retaining walls with proper drainage, soldier piles, soil nails, horizontal drains, and subsurface dewatering all work — and all require a licensed geotechnical engineer, because an undrained retaining wall is a dam that will fail. Budget in the tens of thousands and up.
  • Learn the warning signs and act on them. New cracks in foundations, slabs, driveways, or bare soil; doors and windows that suddenly stick; tilting trees, fence posts, or utility poles; new springs, seeps, or unusually wet ground; bulging at the base of a slope; and sagging or taut utility lines. These are how a slow failure announces itself, sometimes months ahead. They are a call to a geotechnical engineer, not something to watch.
  • Below a burn scar, treat every debris flow warning as immediate evacuation. The NWS and USGS issue targeted post-fire debris flow warnings, and the flows arrive within minutes of the rainfall. There is no version of this where you leave after it starts.

Insurance: the biggest gap in this index

Standard homeowners policies exclude earth movement — landslide, mudflow, earthflow, subsidence, and sinkholes alike. This is one of the widest coverage gaps in American property insurance, and most homeowners in high-susceptibility areas do not know it exists until they file.

  • Difference in Conditions (DIC) policies are the main route to coverage, sold in the surplus-lines market, typically bundling earthquake and landslide. They are expensive, underwritten property by property, and often unavailable on a slope with visible distress — which is to say, availability itself is a risk signal worth listening to.
  • NFIP flood insurance covers “mudflow” — a river of liquid mud flowing on the surface — but not landslide or earth movement generally. The distinction is technical and consequential: post-fire debris flows sometimes qualify as mudflow and sometimes do not, and it is decided case by case. If you live below a burn scar, a flood policy is worth having and is not a guarantee.
  • Sinkhole coverage is a separate market in Florida and a handful of other states with karst geology, sometimes mandated as an offer.
  • FEMA individual assistance after a declared disaster is capped and designed for immediate needs, not for rebuilding a house or stabilizing a slope. Do not plan around it.
  • Liability runs uphill and downhill. Altering drainage or grading in a way that damages a neighbor’s property creates real exposure, and it works both ways. Document the condition of your slope with dated photographs each year.

Before, during, and after

Before. Find out what your state publishes — Washington, Oregon, California, Colorado, Utah, and North Carolina all maintain landslide susceptibility or inventory mapping, much of it now lidar-derived and far better than what existed a decade ago — and look at the USGS Landslide Hazards Program national layers. Before buying in hill country, ask specifically about slide history in the disclosures, look for repaired cracks, patched driveways, and retaining walls added after construction, and pay for a geotechnical review if anything on the lot suggests movement. Know whether you are on, above, or below a slope — being below is often the more dangerous position, and the one people fail to consider.

During. In intense rain in steep terrain, stay awake and alert. Listen: debris flows announce themselves with cracking trees, knocking boulders, and a rumble that grows quickly. Water in a stream that suddenly turns muddy, or a stream that suddenly stops flowing (an upstream blockage), are both immediate evacuation signals. Move out of the path laterally and upslope, not down-channel — you cannot outrun a flow moving at highway speeds down its own channel. Upper floors are better than lower ones if you cannot get out. Canyon mouths, alluvial fans, and channel bottoms are the kill zones.

After. Stay out of the slide area: secondary failures are common in the hours and days afterward, and the ground is unstable in ways that are not visible. Watch for flooding from blocked drainages. Report broken utility lines. Get a geotechnical assessment before any repair or re-entry to a structure on or below the failure, and document everything photographically before cleanup — with earth movement excluded from most policies, the paper trail matters for FEMA assistance, litigation, or a DIC claim. Assume the hazard has not gone away: slide sites repeat, and the same slope that moved once is now weaker than it was.

Sources

See the US counties with the highest landslide risk →

← All guides