What are the odds of dying in a landslide?
Evidence quality 4.0/5
Eight-dimension review score against the quality rubric . Each dimension scored 1–5.
- D1 Source grounding
- 4/5
- D2 Source authority
- 5/5
- D3 Arithmetic
- 5/5
- D4 Uncertainty
- 3/5
- D5 Scope
- 4/5
- D6 Prose
- 4/5
- D7 Perception honesty
- 3/5
- D8 Caveat completeness
- 4/5
≈ As likely as
Perceived
Landslides rarely feature in the mental inventory of natural-disaster threats. Earthquakes, hurricanes, and floods dominate the fear hierarchy; landslides are perceived as a secondary consequence rather than an independent killer. This is partly a classification artifact — many landslide deaths are attributed to the earthquake or storm that triggered them — and partly a visibility problem: landslides tend to kill in small, dispersed events across mountainous terrain in low-income countries, generating little international media coverage per incident.
Rough estimate: most people would rank landslides well below earthquakes, floods, and storms as a cause of death — the actual annual toll is comparable
Source: editorial intuition, not polled
Actual
~4,600 deaths per year globally (2004–2016 average, Froude & Petley 2018)
global adults
Show derivation
Froude and Petley (2018) documented 55,997 fatalities in 4,862 distinct landslide events from 2004 to 2016 using the Global Fatal Landslide Database — an average of ~4,615 deaths per year. This figure is itself an undercount: Petley (2012) showed that the widely used EM-DAT disaster database undercounted landslide fatalities by over 400%. Annual rate: 4,600 / 5,000,000,000 = 9.2 × 10⁻⁷. Compounded over 59 years: 1 − (1 − 9.2e-7)^59 ≈ 5.43 × 10⁻⁵, i.e. roughly 1 in 18,400. The uncertainty band uses a low of ~3,000 deaths/year (low: 3.54e-5) and a high of ~6,500/year (high: 7.67e-5).
Caveats: Risk is heavily concentrated in mountainous regions of South and Southeast Asia …
Risk is heavily concentrated in mountainous regions of South and Southeast Asia (India, Nepal, China, Philippines, Indonesia), Central and South America (Colombia, Brazil, Guatemala), and parts of sub-Saharan Africa. Steep terrain, heavy rainfall, deforestation, and informal construction on unstable slopes are the primary risk factors. Residents of flat, geologically stable terrain face effectively zero risk. The 4,600 deaths/year figure excludes seismically triggered landslides, which would add significantly to the total in earthquake-prone years. Climate change is expected to increase landslide frequency through more intense rainfall events, though the magnitude of the effect is uncertain. A distinct, time-limited vector applies to households below a recent wildfire burn scar: the USGS reports that post-fire debris flows can occur in the years immediately after wildfires in response to high-intensity rainfall, with debris-flow activity in recently burned areas typically occurring within two years of the fire. These slurry flows can move with little warning when short, intense rainfall exceeds USGS-modelled basin thresholds; the January 2018 Montecito flow on the Thomas Fire burn scar killed 23 people. The size of the personal-death multiplier for a below-scar resident has not been quantified — USGS publishes per-basin debris-flow likelihoods and rainfall-intensity thresholds rather than a population fold-change — so this entry records the vector without adjusting the headline number.
How the risk varies
The headline figure averages across very different situations. Here’s how the probability varies by scenario or context:
1 in 4,000
India, Nepal, China, Philippines, and Indonesia account for the majority of global landslide fatalities. Monsoon-season rainfall on deforested slopes is the primary trigger.
1 in 8,333
Colombia, Brazil, and Guatemala report significant annual fatalities, often in informal settlements on steep terrain.
1 in 666,667
Approximately 25-50 deaths per year in the US; landslide risk mapping and building codes substantially reduce exposure.
Bar length and shade rank these scenarios against each other, not against other risks. The exact odds are shown beside each.
Related risks
Other risks on similar themes — for exploring related fears.
Drought famine death
What are the odds of dying from drought-induced famine or water scarcity?
Post-hurricane heat death
How likely is dying from heat after a hurricane knocks out your power in summer?
Pick challenger
Landslides kill roughly 4,600 people every year — a figure established by Froude and Petley’s 2018 analysis of the Global Fatal Landslide Database, which tracked 55,997 deaths across 4,862 events from 2004 to 2016. That number is almost certainly an undercount. Standard disaster databases like EM-DAT attribute landslide deaths to the earthquake or rainstorm that triggered them, a classification habit that Petley showed undercounts landslide fatalities by over 400%. The USGS independently corroborates the ~4,500/year figure. At this rate, the average global adult lifetime probability is about 1 in 18,400 — roughly comparable to the lifetime risk of dying in a flood.
The perception deficit is largely a function of event size and media geography. Landslides tend to kill in small, dispersed incidents — a village buried here, a road crew struck there — rather than in single headline-generating catastrophes. The deadliest events occur in South and Southeast Asia during monsoon seasons, where steep terrain, heavy rainfall, deforestation, and informal construction on unstable hillsides create a lethal combination. India, Nepal, China, the Philippines, and Indonesia collectively account for the majority of global landslide deaths. When a landslide does make international news — Oso, Washington in 2014 (43 dead); Sierra Leone in 2017 (over 1,000 dead) — the reaction is surprise that something so mundane could be so deadly.
Climate change is expected to worsen the problem. More intense rainfall events increase pore-water pressure in slopes, triggering more frequent and larger failures. Deforestation for agriculture and development removes root systems that stabilize hillsides. Urbanization pushes the poorest populations onto the steepest, cheapest, most landslide-prone terrain. The result is a risk that is growing in magnitude but shrinking in visibility — exactly the combination that makes a hazard easy to ignore until it arrives.
Related tidbits
Landslides kill about 4,600 people a year globally, roughly 1 in 18,400 over a global adult lifetime. Most rank them well below earthquakes, floods, and storms, yet the annual toll is comparable.
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
-
[1] Natural Hazards and Earth System Sciences (Copernicus) — Global fatal landslide occurrence from 2004 to 2016
Global fatal landslide occurrence from 2004 to 2016- Statistic
55,997 people were killed in 4,862 distinct non-seismically triggered landslide events between 2004 and 2016, averaging ~4,615 deaths per year- Excerpt
“"In total, 55,997 people were killed in 4,862 distinct landslide events. The data show that the number of recorded fatal landslides and resulting fatalities has increased over the study period, with particularly high counts in South and Southeast Asia." ”
- Source data from
- 2018-08-01
- Accessed
- 2026-04-24 · archived copy
- Calculation
- Froude and Petley's 13-year dataset from the Global Fatal Landslide Database is the most comprehensive peer-reviewed accounting of global landslide mortality. At 4,600 deaths/year over a global adult population of 5 billion, the annual rate is 9.2e-7; compounded over 59 years: ~5.43e-5. The EM-DAT database, by contrast, would yield a figure several hundred percent lower.
- Independence
- Independent academic dataset (Durham/Sheffield Global Fatal Landslide Database), methodologically separate from the USGS reference below.
-
[2] U.S. Geological Survey — How many deaths result from landslides each year?
How many deaths result from landslides each year?- Statistic
25-50 people killed by landslides each year in the US; worldwide death toll is in the thousands- Excerpt
“"An average of 25-50 people are killed by landslides each year in the United States. The worldwide death toll per year due to landslides is in the thousands." ”
- Source data from
- 2024-01-01
- Accessed
- 2026-04-24 · archived copy
- Calculation
- The USGS confirms "worldwide death toll per year due to landslides is in the thousands," which is consistent with the Froude & Petley estimate of ~4,600/year. The USGS also provides the US-specific figure of 25-50 deaths per year, which anchors the US regional breakdown entry below.
- Independence
- USGS is a U.S. federal science agency; its landslide mortality estimates draw on multiple international sources independently of the Sheffield/Durham academic database.
-
[3] U.S. Geological Survey — What should I know about wildfires and debris flows?
What should I know about wildfires and debris flows?- Statistic
Post-fire debris flows can occur in the years immediately after wildfires in response to high-intensity rainfall; debris-flow activity in recently burned areas typically occurs within 2 years of the fire.- Excerpt
“"Post-fire landslide hazards include fast-moving, highly destructive debris flows that can occur in the years immediately after wildfires in response to high intensity rainfall events." ”
- Source data from
- 2024-01-01
- Accessed
- 2026-06-14
- Calculation
- Cited only to substantiate the post-fire burn-scar debris-flow caveat, not the headline rate. The USGS Hazard Assessment FAQ separately states that "debris-flow activity in recently burned areas typically occurs within 2 yr of wildfire," establishing the elevated-risk time window. USGS publishes per-basin debris-flow likelihoods and rainfall-intensity thresholds rather than a population-level fold-change, so no personal multiplier is derived. The January 2018 Montecito event on the Thomas Fire burn scar (23 fatalities, widely reported) is the canonical illustration of this conditional hazard.
- Independence
- Same agency (USGS) as the landslide-mortality FAQ above, but a distinct hazard product (postfire debris-flow program); used for the burn-scar caveat, while the global headline rests on the independent Froude & Petley academic dataset.