Apartment fire statistics: multifamily fires reported to NFIRS, 2020 to 2025

This month in brief

• Departments reported 60,458 multifamily building fires to NFIRS in 2024 and 69,359 in 2023; the 2024 dip comes from the largest reporter's mostly missing 2024 file.
• In a panel of 2,026 departments that reported every year, fires rose 11.4% from 2020 to 2024 and nonconfined fires rose 16.3%.
• Civilian deaths per 1,000 nonconfined fires fell from 10.05 in 2020 to 8.26 in 2024.
• Cooking started 60.4% of fires in 2024 (largest reporter excluded); smoking started 1.8% of fires but caused 9.0% of deaths from 2020 to 2024.
• 29.9% of nonconfined fires spread beyond the room of origin, falling from 34.9% in one-story buildings to 12.4% in 7 to 12 story buildings.
• With sprinklers or other AES present, 12.1% of fires spread beyond the room, against 33.6% with none.
• High-rise fires injured 130.5 civilians per 1,000 fires, against 58.1 in low-rise buildings.
• NFIRS public data has no construction-type field, so fire-rated construction cannot be tested; 2025 data are partial because of the switch to NERIS.

Apartment fire statistics by year, 2020 to 2025

Raw NFIRS counts swing with reporting coverage, so the yearly tracker shows three series. The consistent panel of 2,026 departments shows fires rising 11.4% from 2020 to 2024, while the raw count fell 12.8% in 2024 because the largest reporter's 2024 file is mostly missing.

Multifamily building fire (this report): A fire reported to NFIRS with property use 429 (multifamily dwelling: apartments, condos, town houses, row houses, tenements), incident type 111 or 120 to 123 in an enclosed or portable building, or a confined fire (incident types 113 to 118). Mutual and automatic aid given records are excluded, matching the USFA multifamily topical report specification.

Departments reported 64,408 multifamily building fires to NFIRS in 2020, 69,359 in 2023 and 60,458 in 2024 (NFIRS Public Data Release).

The 2024 dip is a reporting artifact. All reported fires fell by 8,901 from 2023 to 2024, and the largest reporter, New York City's fire department (NFIRS ID NY 24001), accounts for a drop of 8,961 records on its own: 11,953 records in 2023 against 2,992 in 2024, or 25.0% of its 2023 volume.

Civilian deaths in reported multifamily building fires were 178 in 2020, 206 in 2023 and 167 in 2024. The death rate per 1,000 nonconfined fires moved from 10.05 in 2020 to 8.26 in 2024, a change of -17.8%.

Civilian injuries per 1,000 nonconfined fires fell from 70.9 in 2020 to 48.5 in 2024 (-31.6%). Firefighter injuries on primary-department reports were 982 in 2020 and 918 in 2024, or 15.2 per 1,000 fires in 2024.

Reported property and contents losses in multifamily building fires reached $1.40 billion in 2024, up 38.3% from $1.01 billion in 2020 after adjusting to 2024 dollars with the BLS CPI-U. The mean reported loss per nonconfined fire rose from $60,936 to $73,940 (21.3%).

The 2025 file is partial. USFA's release notes say 21,366 departments reported in 2025 against 22,582 in 2024 (-5.4%), partly because NFIRS was decommissioned in February 2026 and replaced by NERIS. In the consistent panel, December 2025 fires were only 0.569 times the December 2024 level, so 2025 counts are shown but kept out of the trend.

Causes of apartment fires: cooking leads counts, smoking leads deaths per fire

Cooking started 63.8% of multifamily building fires from 2020 to 2024 but only 3.1% of deaths. Smoking started 1.8% of fires and 9.0% of deaths, the highest death rate of any identified cause.

USFA main cause: NFIRS assigns each structure fire one of 16 main cause categories (the CAUSE_CODE field in causes.txt) through a 38-step hierarchy: exposure first, then intentional, cause under investigation, playing with heat source, natural, and so on down to unknown. A fire falls in the highest category it qualifies for (USFA Structure Fire Cause documentation). PCC is the 35-value priority cause code and GCC the 7-value general cause code.

Excluding the largest reporter, cooking's share of multifamily building fires moved from 63.8% in 2020 to 60.4% in 2024 (-3.4 percentage points). Unknown cause rose from 11.9% to 14.1% (2.2 points).

Heating held at 2.5% of fires in 2024, electrical malfunction rose from 2.3% to 2.7%, smoking fell from 2.2% to 1.7%, and intentional fires fell from 2.5% to 2.3%.

From 2020 to 2024, smoking fires killed 15.14 civilians per 1,000 fires, against 0.14 for cooking fires. Smoking's share of deaths was 5.0 times its share of fires. Intentional fires caused 6.6% of deaths at 8.95 per 1,000 fires.

Unknown-cause fires made up 12.8% of fires but 47.5% of deaths, because fatal fires are harder to reconstruct. That gap caps how precisely any cause ranking for deaths can be stated. NFIRS cause codes do not single out lithium-ion batteries; see the lithium-ion battery fires report for recall and city data.

Where in the building nonconfined fires begin

Kitchens were the area of origin in 28.8% of nonconfined multifamily fires from 2020 to 2024, but bedrooms and living rooms account for more of the deaths.

Bedrooms were the origin of 10.8% of nonconfined multifamily building fires and 21.4% of civilian deaths from 2020 to 2024. Living rooms and other common rooms were the origin of 6.7% of fires and 23.8% of deaths.

Exterior balconies and unenclosed porches started 6.0% of nonconfined fires, and 46.2% of those spread beyond the area of origin. Kitchen fires spread beyond the room in 12.7% of cases, and attic fires in 61.8%.

How far fires spread, by building height

The share of nonconfined fires that spread beyond the room of origin falls with building height, from 34.9% in one-story buildings to 12.4% in 7 to 12 story buildings.

Across all 95,835 nonconfined multifamily building fires reported in 2020 to 2024, 29.9% spread beyond the room of origin and 19.5% spread beyond the floor of origin. The fire spread field had no blanks for nonconfined fires.

Buildings of 4 to 6 stories saw 18.7% of fires spread beyond the room, and buildings of 13 or more stories 18.0%. Sprinkler presence rises with height, from 8.4% of known cases in two-story buildings to 55.1% in 7 to 12 story buildings, which explains part of the gradient.

Building height was blank, zero or above 110 stories for 1,165 nonconfined fires (1.2%), which are shown separately.

Sprinklers present or absent: spread and deaths in apartment fires

In nonconfined multifamily fires where an automatic extinguishing system was present, 12.1% spread beyond the room, against 33.6% where none was present. The gap holds within both low-rise and high-rise buildings.

From 2020 to 2024, fires in buildings with no automatic extinguishing system spread beyond the room 2.8 times as often as fires where one was present. Civilian deaths were 11.9 per 1,000 fires without a system and 3.1 per 1,000 with one (3.8 times), based on 50 deaths in 16,094 fires with a system present.

In low-rise buildings, 34.1% of fires without a system spread beyond the room, against 12.3% with one (2.8 times). In high-rise buildings, the shares were 18.7% and 10.1% (1.9 times).

Partial systems were recorded in 568 nonconfined fires, with 22.7% spreading beyond the room. Presence was undetermined or blank in 7,656 fires.

High-rise and low-rise apartment fires compared

High-rise multifamily fires (7 or more stories) spread beyond the room less often, 14.7% against 30.7%, but injured 2.2 times as many civilians per fire.

High-rise buildings accounted for 4,805 of the 95,835 nonconfined multifamily fires reported from 2020 to 2024 (5.0%). The 7-story threshold follows the NFPA high-rise analysis convention of seven stories or more.

Civilian deaths per 1,000 fires were similar: 9.4 in high-rise buildings and 9.9 in low-rise buildings. Civilian injuries were 130.5 per 1,000 high-rise fires against 58.1 in low-rise buildings, and firefighter injuries 76.0 against 43.1 (1.8 times).

A system was present (full or partial) in 50.9% of high-rise fires with a known status, against 17.2% in low-rise buildings. The median reported loss per nonconfined fire was $1,000 in high-rise buildings and $2,894 in low-rise buildings (2024 dollars).

Deaths in high-rise fires were few: 45 in five years, 29 of them in buildings with no system present and 15 with a system present. Rates built on counts this small move a lot from year to year.

Smoke alarm presence in reported apartment fires

No smoke alarm was present in 15.7% of nonconfined multifamily fires in 2024, or 19.3% of fires where presence was known. The undetermined share grew from 16.6% to 18.6%.

Smoke alarms were recorded as present in 65.7% of nonconfined multifamily building fires in 2024 and 66.4% in 2020. USFA reported 64% for 2017 to 2019 in its topical report.

Pooled over 2020 to 2024, no alarm was present in 20.3% of nonconfined fires with known status. Civilian deaths per 1,000 fires were 8.8 with no alarm present and 9.7 with an alarm present, which shows that presence alone says little: whether the alarm operated and alerted occupants matters more, and that is outside this report's scope.

Alarm requirements for apartments vary by state; see the smoke detector laws by state report and the fire alarm inspection page for building systems.

Monthly seasonality of multifamily fires

Multifamily fires run about 21.8% hotter per day in the peak month than in the quietest month, driven by heating and cooking in winter.

From 2020 to 2024, excluding the largest reporter, multifamily building fires averaged 160.0 per day in Feb and 131.4 per day in Aug, the lowest month. On an index where the average month is 100, Feb scored 107.9 and Aug 88.6.

December to February accounted for 40.1% of heating fires, and January to March for 34.6% of civilian deaths. Cooking fires were steadier, between 12,141 and 15,180 per month across the five pooled years.

Fire-rated construction: a field NFIRS does not carry

The question of how fire-rated construction relates to deaths and losses cannot be answered from NFIRS 5.0 public data, because the release has no construction-type field.

The NFIRS 5.0 structure fire module records structure type, building status, height, main floor size, fire origin, fire spread, detectors and automatic extinguishing systems (NFIRS Complete Reference Guide). It does not record construction type or fire-resistance rating, and the Public Data Release Format v1.8 has no such column.

STRUC_TYPE, which the brief listed, means enclosed building, portable structure, tent and so on, not construction class. Building height and sprinkler presence are the closest proxies in the data, and both are reported above. Property managers looking for the code side of construction and compartmentation can start at the apartments and HOA buildings guide.

Momentum and share analysis

Download the calculations (CSV)

Findings to track

1. Fire departments reported 60,458 multifamily building fires to NFIRS in 2024, 66.6% of them confined fires.^[6]
2. Among 2,026 departments that reported multifamily fires in every year, fires rose from 39,841 in 2020 to 44,366 in 2024 (11.4%).^{[2][3][4][5][6]}
3. The largest reporter's multifamily records fell by 8,961 from 2023 to 2024, which accounts for the raw 2024 decline of 8,901 fires.^[5][6]
4. Civilian deaths in reported multifamily building fires totaled 943 from 2020 to 2024, with 167 in 2024.^{[2][3][4][5][6]}
5. Civilian deaths per 1,000 nonconfined multifamily fires moved from 10.05 in 2020 to 8.26 in 2024.^[2][6]
6. Cooking caused 63.8% of multifamily building fires from 2020 to 2024 and 3.1% of civilian deaths.^{[2][3][4][5][6][8]}
7. Smoking caused 1.8% of multifamily building fires from 2020 to 2024 and 9.0% of civilian deaths, 15.14 deaths per 1,000 fires.^{[2][3][4][5][6][8]}
8. Electrical malfunction rose from 2.3% of multifamily fires in 2020 to 2.7% in 2024, excluding the largest reporter.^{[2][3][4][5][6][8]}
9. Bedrooms were the origin of 10.8% of nonconfined multifamily fires but 21.4% of deaths in 2020 to 2024.^{[2][3][4][5][6]}
10. 29.9% of nonconfined multifamily fires in 2020 to 2024 spread beyond the room of origin, against 31% reported by USFA for 2017 to 2019.^{[2][3][4][5][6][11]}
11. Fires spread beyond the room in 12.1% of nonconfined multifamily fires with an automatic extinguishing system present and 33.6% with none, 2020 to 2024.^{[2][3][4][5][6]}
12. High-rise multifamily fires (7+ stories) spread beyond the room in 14.7% of cases, against 30.7% in low-rise buildings, 2020 to 2024.^{[2][3][4][5][6][18]}
13. Firefighters were injured at 76.0 per 1,000 high-rise multifamily fires and 43.1 per 1,000 low-rise fires, 2020 to 2024.^{[2][3][4][5][6]}
14. No smoke alarm was present in 19.3% of 2024 nonconfined multifamily fires with known alarm status.^[6]
15. Reported losses in multifamily building fires were $1.40 billion in 2024, up 38.3% from 2020 in 2024 dollars.^{[2][3][4][5][6][15]}

Frequently asked questions

Related guides, services and research

• Apartment and HOA fire safety requirements
• Fire alarm inspection requirements
• Fire sprinkler inspection requirements
• Carbon Monoxide Detector Laws by State
• Fire Alarm False Alarm Fines in the 50 Largest U.S. Cities (2026)
• Fire Alarm System Installation Cost per Square Foot
• Fire Alarm License Requirements by State

Methodology

Selection: NFIRS Public Data Release fire and hazmat files for 2020 to 2025, queried from a read-only DuckDB copy. The universe follows USFA Topical Fire Report v21i7: property use 429, aid types 3 and 4 excluded, incident types 111 and 120 to 123 with structure type 1 or 2, and confined incident types 113 to 118 with structure type 1, 2 or blank. Incident type 112 is excluded, as USFA has done since 2008 data.

Property use 439 (boarding and rooming houses, residential hotels) is excluded because USFA's analysis guidelines classify it as other residential, not multifamily, and its occupancy (single rooms, shared facilities) differs. A side count is in tables/t1c_boarding_rooming_439_side_count.csv.

Exclusions: incidents listed as exclude-from-all-analyses in each year's Data Quality Errata Addendum (including 45 duplicate pairs in 2025 and test incidents in 2023) and any duplicated INCIDENT_KEY. Dollar-loss errata and placeholder values (999,999 to 999,999,999) are excluded from loss figures only. Civilian deaths from OTH_DEATH match civiliancasualty severity-5 records in every year.

Coding: every code was confirmed in codelookup. FIRE_SPRD 1 object, 2 room, 3 floor, 4 building, 5 beyond building; beyond room = 3 to 5. AES_PRES 1 or Y present, 2 partial, N none, U undetermined. DETECTOR 1 or Y present, N none, U undetermined. Causes use CAUSE_CODE (USFA main fire cause, 16 values) from causes.txt; PCC is the priority cause code and GCC the general cause code. Height uses BLDG_ABOVE (stories at or above grade); 0 and values above 110 are treated as invalid; high-rise means 7 or more stories.

Spread, height, sprinkler and smoke alarm analyses use nonconfined fires only, because NFIRS does not require the structure fire module for confined fires (USFA analysis guidelines). Shares show unknown and undetermined categories; where a share excludes them, the column says (known).

Trend handling: NFIRS is voluntary and participation changes. The largest reporter (NY 24001) has fewer than 100 multifamily records in each of nine months of 2024, so trend views show raw counts, counts excluding it, and a consistent panel of 2,026 departments with records in all six years. Cause shares by year and monthly seasonality exclude it. 2025 is partial (NFIRS decommissioned February 2026) and is shown but not used in trends. States are not ranked.

Firefighter injuries: USFA guidance says aid-given records should be included when counting firefighter casualties. Primary-department records are shown in the main table; aid-given records with property use 429 are listed separately in tables/t1_yearly_tracker.csv. 2025 includes an unusual number of incidents with 10 or more firefighter injuries (tables/t1g_firefighter_injury_outliers.csv).

Losses: property plus contents loss as reported, converted to 2024 dollars with BLS CPI-U annual averages. Medians include zero-loss reports. Losses are not reported for every fire; 92.6% of nonconfined fires carried a loss value.

Comparisons to USFA's 2017 to 2019 topical report are directional: USFA figures are national estimates, ours are reported counts and shares. Construction type is not in NFIRS 5.0 public data, so fire-rated construction is not analyzed. Last updated 2026-11-01. Scripts: scripts/errata_keys.py, build_universe.py, compute.py, supplement.py, build_report.py.

This report presents published data with its limitations. It is not legal, engineering, financial or investment advice. Requirements depend on the code edition your state and local fire marshal have adopted; confirm with your authority having jurisdiction.

Sources, ranked by quality

Tier 1: primary and official sources (17)

• [1] FEMA OpenFEMA / U.S. Fire Administration. FEMA/USFA NFIRS Annual Data (dataset page, updated June 25, 2026). 2004 to 2025 fire and hazmat files. Accessed 2026-11-01.
• [2] U.S. Fire Administration (FEMA). NFIRS 2020 Public Data Release, fire and hazmat file. 2020. Accessed 2026-11-01.
• [3] U.S. Fire Administration (FEMA). NFIRS 2021 Public Data Release, fire and hazmat file. 2021. Accessed 2026-11-01.
• [4] U.S. Fire Administration (FEMA). NFIRS 2022 Public Data Release, fire and hazmat file. 2022. Accessed 2026-11-01.
• [5] U.S. Fire Administration (FEMA). NFIRS 2023 Public Data Release, fire and hazmat file. 2023. Accessed 2026-11-01.
• [6] U.S. Fire Administration (FEMA). NFIRS 2024 Public Data Release, fire and hazmat file. 2024. Accessed 2026-11-01.
• [7] U.S. Fire Administration (FEMA). NFIRS 2025 Public Data Release, fire and hazmat file. 2025 (released 2026-05-01; partial participation, NERIS transition). Accessed 2026-11-01.
• [8] U.S. Fire Administration. Cause for Structure Fires (Structure Fire Cause.pdf, NFIRS PDR documentation). Cause hierarchy applied to 2020 to 2025 PDRs. Accessed 2026-11-01.
• [9] U.S. Fire Administration. NFIRS 5.0 Fire Data Analysis Guidelines and Issues. Methodology. Accessed 2026-11-01.
• [10] U.S. Fire Administration. NFIRS 5.0 Complete Reference Guide (2015). Field definitions. Accessed 2026-11-01.
• [11] U.S. Fire Administration. Multifamily Residential Building Fires (2017-2019), Topical Fire Report Series Vol. 21 Issue 7. 2017 to 2019 (published August 2021). Accessed 2026-11-01.
• [12] U.S. Fire Administration. Residential Fire Estimate Summaries (2014-2023). 2014 to 2023 national estimates. Accessed 2026-11-01.
• [13] U.S. Fire Administration. National Fire Incident Reporting System (program page, 2025 reporting totals as of Jan. 5, 2026). 2025 year to date. Accessed 2026-11-01.
• [14] U.S. Fire Administration (operated with UL FSRI). NERIS: The National Emergency Response Information System. Current. Accessed 2026-11-01.
• [15] U.S. Bureau of Labor Statistics. Consumer Price Index for All Urban Consumers (CUUR0000SA0), annual averages. 2020 to 2025. Accessed 2026-11-01.
• [16] U.S. Fire Administration. NFIRS Data Quality Errata Addenda, PDR 2020 to 2025. 2020 to 2025. Accessed 2026-11-01.
• [17] U.S. Fire Administration. NFIRS Public Data Release Format, Version 1.8. File layout (dated 9/24/2020). Accessed 2026-11-01.

Tier 2: trade bodies, market research, company data (1)

• [18] National Fire Protection Association (copy hosted by ICC). Reported Fires in High-Rise Structures in Selected Occupancies with and without Automatic Extinguishing Systems (M. Ahrens, December 2010). Definition of high-rise (7+ stories). Accessed 2026-11-01.

Tier 3: journalism and expert commentary (1)

• [19] Verisk. Fire trends in multifamily housing (December 1, 2016). Summary of USFA 2012 to 2014 report. Accessed 2026-11-01.