Holiday fire statistics: a day-by-day index of fires reported to NFIRS, 2020 to 2024

This month in brief

• The four days with the most fires reported to NFIRS in 2020 to 2024 were July 4, July 5, July 3 and January 1: all four are fireworks days.
• The Fourth of July averaged 10,473 fires against a baseline of 2,616, an index of 4.00 and about 7,857 extra fires per holiday.
• Thanksgiving is a cooking holiday, not a fire holiday in general: cooking fires ran 3.57 times the baseline while all fires ran 1.22 times and outside fires 0.90 times.
• Christmas Eve and Christmas Day add cooking fires (1.69 and 1.77 times), candle fires (2.43 and 1.90 times) and decoration fires (4.54 and 4.12 times), but all fires rise only 12% and 8%.
• NFIRS holds 634 fires with a Christmas tree as the item first ignited in five years; 324 were structure fires, and those carried 1.54 civilian deaths per 100 fires against 0.94 for all residential structure fires.
• Civilian deaths did not spike with fire counts: the death index ranged from 0.87 on Christmas Eve to 1.21 on New Year's Day.
• The pattern repeats every year: the July 4 all-fire index stayed between 3.65 and 4.28, and the Thanksgiving cooking index between 2.80 and 3.97.

Holiday fire statistics across the calendar year: which days have the most fires

Raw counts and weekday-adjusted ratios point to the same answer: fireworks days lead the year. July 4, July 5, July 3 and January 1 are the four days with the most fires reported to NFIRS in 2020 to 2024, and they also hold four of the top four places on the index.

Holiday fire index: Fires reported to NFIRS on a holiday divided by a baseline: the mean of the same weekday 7, 14, 21 and 28 days before and after, with other holidays removed. An index of 1.00 is a normal day for that weekday and season. Pooled index = five holidays (2020 to 2024) added together, divided by their five baselines added together.

The Fourth of July stands apart. It averaged 10,473 fires against a baseline of 2,616, which works out to about 7,857 extra fires every year. New Year's Day is second at 1.64 times, then Thanksgiving at 1.22, New Year's Eve at 1.18, Christmas Eve at 1.12 and Christmas Day at 1.08.

Raw counts mislead outside the holidays. Late March and early April fill places 5 to 8 and 10 to 12 on the raw ranking because spring is the peak season for outdoor and debris fires, yet their index sits near 1.00 (for example April 3 at 1.05). The weekday-matched baseline is what separates a holiday effect from a seasonal one.

Civilian deaths do not follow the fire counts. The death index was 1.05 on July 4, 0.99 on Thanksgiving, 1.05 on Christmas Day, 0.87 on Christmas Eve, 0.88 on New Year's Eve and 1.21 on New Year's Day. Many of the extra holiday fires are outdoor or confined cooking fires, which rarely injure anyone. State-level death rates are in our report on fire deaths by state.

January 1: New Year's Day fires, fireworks and discarded trees

New Year's Day is the second-biggest holiday for fires in NFIRS, driven by outdoor and fireworks fires around midnight, and it is also the peak day for Christmas tree fires.

New Year's Day averaged 3,574 fires in 2020 to 2024 against a baseline of 2,181, an index of 1.64. Outside fires ran 2.18 times the baseline and structure fires 1.35 times. Cooking fires rose less (1.18 times).

Fireworks fires averaged 230.0 on January 1 against 2.4 on a baseline day. New Year's Eve averaged 134.4. Together, December 31 and January 1 held 1,822 fireworks fires in five years, 7.6% of the annual total.

Christmas tree fires peak after Christmas, not before. January 1 averaged 6.4 tree fires against 1.4 on baseline days (4.48 times), and January holds 278 tree fires in five years against 162 in December. Most January tree fires are outside or rubbish fires: 162 of them, 65.4% coded intentional. The coding fits discarded trees being burned rather than trees igniting indoors.

Fourth of July fire statistics: the busiest fire day in NFIRS

July 4 is four times a normal day for all fires, close to six times for outside fires, and the fireworks effect spills into July 3 and July 5.

Across 2020 to 2024, July 4 averaged 10,473 fires reported to NFIRS. Outside fires averaged 7,523 (5.77 times baseline), structure fires 2,073.2 (2.58 times) and cooking fires 317.0 (1.22 times).

The effect spreads over three days. July 5 averaged 6,043 fires (index 2.31) and July 3 averaged 3,862 (1.48). Of 23,987 fireworks-ignited fires reported to NFIRS in five years, 12,656 (52.8%) fell on July 3, 4 or 5, and 8,214 (34.2%) on July 4 alone. Most fireworks fires are outdoors: 73.6% were outside fires and 17.9% structure fires.

Fireworks are a floor, not a full count. The heat source is unknown on 41.8% of July 4 fire module records, and many outdoor fires have no fire module at all. Where the heat source is known, fireworks account for 49.3% of July 4 fires.

Thanksgiving cooking fires statistics: 3.57 times a normal Thursday

Thanksgiving is the strongest cooking-fire day of the year in NFIRS. The rise is concentrated in late morning and early afternoon, while outside fires fall below a normal Thursday.

Thanksgiving (the fourth Thursday of November) averaged 941 cooking fires in 2020 to 2024 against 263.2 on baseline Thursdays, an index of 3.57. That is 677.8 extra cooking fires per Thanksgiving. Structure fires ran 1.81 times the baseline, all fires 1.22 times and outside fires 0.90 times.

Cooking fires here are structure fires that the USFA cause hierarchy assigns to cooking, and 94.1% of them are confined cooking fires (incident type 113). On Thanksgiving, cooking caused 61.2% of structure fires, against 31.7% across all days.

The hourly curve shows when the extra fires happen. Thanksgiving cooking fires reach 94.0 at the 12:00 hour against 15.2 on baseline Thursdays (6.19 times), and the hourly index peaks at 6.46 at 14:00. 62.4% of Thanksgiving cooking fires were reported between 10:00 and 16:59, against 39.8% on a normal Thursday, when cooking fires cluster around dinner.

Other measures agree on the direction. A USFA snapshot for 2017 to 2019 put Thanksgiving residential building fires at 2.3 times the average of all other days, with cooking behind 73.5% of those with a known cause. NFPA's estimate of about 1,446 home cooking fires on Thanksgiving 2023 (388% of the daily average) is reported here second-hand, through Denver7. Our index differs because it compares Thanksgiving only with nearby Thursdays.

Cooking-cause structure fires include commercial kitchens as well as homes; our data on restaurants covers the commercial side, and the inspection cycle for portable units is in fire extinguisher inspections.

Christmas Eve and Christmas Day: cooking, candle and decoration fires

Christmas changes what burns more than how much burns. All fires rise only slightly, while cooking, candle and decoration fires climb well above a normal day.

Christmas Eve averaged 2,502 fires against a baseline of 2,241 (1.12 times) and Christmas Day 2,415 against 2,237 (1.08 times). Outside fires were flat or lower (0.93 and 1.00), which offsets indoor increases.

Indoor categories rise sharply. Cooking fires ran 1.69 times the baseline on Christmas Eve and 1.77 times on Christmas Day. Candle fires ran 2.43 and 1.90 times. Decoration fires ran 4.54 and 4.12 times, from small counts (7.8 and 6.2 a day). Structure fires rose 1.37 and 1.28 times.

In multifamily buildings, one unit's holiday fire can reach common areas; our guide to apartment and HOA buildings covers the building systems involved, and fire alarm inspections covers alarm testing schedules.

Christmas tree fire statistics: 634 tree fires in five years

Christmas tree fires are rare in NFIRS, but the ones inside buildings carry higher casualty and loss rates than the average residential structure fire. The ratios are smaller than the old USFA figures often repeated online.

NFIRS recorded 634 fires with a Christmas tree as item first ignited in 2020 to 2024, a mean of 126.8 a year. Of these, 324 were structure fires (64.8 a year) and 310 were outside or other fires. 74.7% of structure tree fires occurred in December or January, against 18.5% of all residential structure fires.

Structure tree fires caused 5 civilian deaths and 16 injuries in five years: 1.54 deaths and 4.94 injuries per 100 fires, against 0.94 and 2.88 for residential structure fires. That is 1.64 times the death rate and 1.71 times the injury rate. Reported loss averaged $84,918 per tree structure fire against $41,270, a ratio of 2.06.

Those ratios are far below the USFA 2002 finding that fatalities per fire were nearly eight times those of other winter holiday fires. The 2002 analysis used older data and a different comparison group, and with 5 deaths in our base a single fire moves the rate.

Among structure tree fires, the heat source was undetermined in 76 cases; arcing was coded in 48 and a candle in 27. State requirements for smoke alarms in homes are compiled in our report on smoke detector laws by state.

Candle and decoration fires by month

December is the peak month for candle and decoration fires, and Halloween shows a smaller decoration spike of its own.

Candle fires averaged 11.05 a day in December and 6.17 a day in July, a ratio of 1.79. December's month index is 1.38 for candles and 1.98 for decorations. NFIRS holds 14,624 candle fires in 2020 to 2024; 82.8% were structure fires.

Halloween is the other decoration day: October 31 ran 2.97 times its baseline for decoration fires and 1.98 times for candle fires, while all fires ran 0.95 times.

Holiday-week calendar: the days on either side

The days next to each holiday show where the effect starts and stops: Thanksgiving Eve already shows more cooking fires, and the July 4 effect lasts into July 5.

The day after Thanksgiving is a quiet day (all fires 0.93 times), and so are December 26 and January 2 (0.97 each). Fireworks indexes on ordinary days such as December 30 (3.33) rest on baselines of a few fires a day and should be read as counts, not trends.

Year-by-year consistency, 2020 to 2024, and why 2025 is left out

The holiday pattern repeats in each of the five years. 2025 is excluded because the final NFIRS year is incomplete.

The July 4 all-fire index ranged from 3.65 (2021) to 4.28 (2024). The Thanksgiving cooking index ranged from 2.80 (2021) to 3.97 (2020). Christmas Eve candle fires vary more (1.40 to 4.32) because daily counts are small.

NFIRS ends with 2025. USFA set Jan. 31, 2026 as the final date for NFIRS reporting, and the 2025 release notes say 21,366 departments reported nearly 1,076,400 fire runs, against 22,582 departments and nearly 1,224,400 fire runs in 2024. December 2025 holds 37,035 fires, 54.0% of December 2024, so 2025 holidays would understate the late-year holidays.

Two known gaps affect individual years: the 2021 release includes only 96 District of Columbia incidents, and the 2024 release has 838,657 New York all-incident records against 1,502,248 in 2023. Because each holiday is compared with nearby days in the same year, gaps like these change counts more than they change the index.

Momentum and share analysis

Download the calculations (CSV)

Findings to track

1. July 4 averaged 10,473 fires reported to NFIRS in 2020 to 2024, 4.00 times its same-weekday baseline of 2,616 and 4.15 times the mean day of the year (2,522 fires).^{[1][2][3][4][5][6]}
2. Outside fires (vegetation, rubbish and other outdoor fires) on July 4 averaged 7,523, which is 5.77 times the baseline; structure fires on July 4 ran 2.58 times the baseline.^{[1][2][3][4][5][6]}
3. Fires with fireworks as the heat source averaged 1,642.8 on July 4 against 14.8 on a baseline day in 2020 to 2024, an index of 111.19.^{[1][2][3][4][5][6]}
4. Of 23,987 fireworks-ignited fires reported to NFIRS in 2020 to 2024, 52.8% occurred on July 3, 4 or 5 and 7.6% on December 31 or January 1.^{[1][2][3][4][5][6]}
5. On July 4, fireworks were the heat source in 49.3% of fires where the heat source was known (8,214 of 16,657, 2020 to 2024).^{[1][2][3][4][5][6]}
6. New Year's Day averaged 3,574 fires in 2020 to 2024, 1.64 times its baseline, with outside fires at 2.18 times; New Year's Eve ran 1.18 times.^{[1][2][3][4][5][6]}
7. Thanksgiving averaged 941 cooking fires against 263.2 on baseline Thursdays in 2020 to 2024, an index of 3.57 and 677.8 extra cooking fires per Thanksgiving.^{[1][2][3][4][5][6]}
8. Cooking caused 61.2% of structure fires reported on Thanksgiving in 2020 to 2024, against 31.7% across all days.^{[1][2][3][4][5][6]}
9. 62.4% of Thanksgiving cooking fires were reported between 10:00 and 16:59, against 39.8% on baseline Thursdays; at the 12:00 hour Thanksgiving averaged 94.0 cooking fires against 15.2.^{[1][2][3][4][5][6]}
10. Christmas Day averaged 487.4 cooking fires in 2020 to 2024 (1.77 times baseline) and Christmas Eve 466.0 (1.69 times).^{[1][2][3][4][5][6]}
11. Candle fires on Christmas Eve ran 2.43 times the baseline and decoration fires 4.54 times, 2020 to 2024.^{[1][2][3][4][5][6]}
12. NFIRS recorded 634 fires with a Christmas tree as item first ignited in 2020 to 2024, a mean of 126.8 a year; 324 were structure fires and 310 were outside or other fires.^{[1][2][3][4][5][6]}
13. Christmas tree structure fires carried 1.54 civilian deaths and 4.94 injuries per 100 fires, against 0.94 and 2.88 for residential structure fires, 2020 to 2024 (5 deaths: a small base).^{[1][2][3][4][5][6]}
14. 52.3% of outside and other Christmas tree fires were coded intentional, and January had 162 of them, 65.4% intentional: a pattern consistent with discarded trees being burned after the holidays.^{[1][2][3][4][5][6]}
15. Candle fires averaged 11.05 a day in December and 6.17 a day in July in 2020 to 2024, a ratio of 1.79.^{[1][2][3][4][5][6]}

Frequently asked questions

Related guides, services and research

• Fire extinguisher inspection
• Fire alarm inspection requirements
• Apartment and HOA fire safety requirements
• Hotel Fire Safety Compliance
• Industrial fires and explosions investigated by the U.S. Chemical Safety Board, 1998 to 2026
• Lithium-Ion Battery Fires Statistics
• Most Common Fire Code Violations

Methodology

Data. We queried the USFA NFIRS Public Data Release (fire and hazardous materials files) for 2020 to 2024: basicincident for incident date, incident type, property use and alarm time; fireincident for heat source (HEAT_SOURC), item first ignited (FIRST_IGN) and cause of ignition; causes for the USFA structure fire cause category; civiliancasualty for civilian deaths (severity 5) and injuries (severity 1 to 4). Records join on INCIDENT_KEY and release year. The universe is fire incidents (incident types 100 to 173), with mutual or automatic aid given (AID 3 or 4) removed to avoid double counting, as USFA guidelines direct. Exposure fires are kept. The result is 4,607,302 fires.

Definitions. Cooking fires = structure fires assigned to the Cooking cause (CAUSE_CODE 05) by the USFA hierarchy, which includes every confined cooking fire (type 113). Outside fires = incident types 140 to 173. Fireworks fires = heat source code 54. Candle fires = heat source code 66. Christmas tree fires = item first ignited code 41. Decoration fires = item first ignited code 42. Residential structure fires = types 111 to 123 with property use 400 to 499. Holidays are computed per year: January 1, July 4, the fourth Thursday of November, December 24, 25 and 31.

Baseline. For each date, the baseline is the mean count on the same weekday 7, 14, 21 and 28 days before and after (eight days), dropping any that fall on a holiday or holiday-adjacent day (January 1, July 3 to 5, October 31, Thanksgiving Wednesday to Friday, December 24, 25 and 31, Easter, Memorial Day, Labor Day) or outside January 1, 2020 to January 31, 2025. January 2025 is used only as baseline days for late-December 2024; holidays on January 1, 2020 use the four days after. Holiday baselines averaged 7.03 usable days. The pooled index adds the five holiday counts and divides by the five baselines; per-year indexes are in t3.

Unknowns. A fire module record exists for 63.8% of fires. Among those, heat source is unknown or blank for 49.7% and item first ignited for 48.5%. Structure fires with USFA cause Unknown are 26.5%. Alarm hour is present on every record. We did not remove unknowns from denominators: fireworks, candle, decoration and tree counts are floors. Christmas tree comparisons use only residential structure fires that also have a fire module, since only those can carry the code.

Limits. NFIRS is voluntary and not a census, so counts are fires reported to NFIRS, not national totals, and we do not rank states. Participation changed between years, which affects counts more than within-year ratios. 2025 is excluded because the final NFIRS year is incomplete (December 2025 holds 54.0% of December 2024's fires). Alarm time is the call time, not the ignition time. Loss values are as reported by departments and not inflation adjusted. This report is statistics, not safety, insurance or legal advice.

Conflicts and other sources. USFA and NFPA publish national estimates that scale NFIRS to all departments; we report their figures only as context and do not mix them with our counts. NFPA report pages could not be read without JavaScript, so NFPA figures appear only as reported by Denver7 (excluded.md). Last updated 2026-11-01. Scripts in scripts/ rebuild every table from the database.

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 (16)

• [1] FEMA / U.S. Fire Administration. FEMA – USFA NFIRS Annual Data (Public Data Release download page). 1980 to 2025 releases. Accessed 2026-11-01.
• [2] U.S. Fire Administration (FEMA). NFIRS 2020 Public Data Release, fire and hazardous materials incidents (basicincident, fireincident, civiliancasualty, causes, codelookup). 2020. Accessed 2026-11-01.
• [3] U.S. Fire Administration (FEMA). NFIRS 2021 Public Data Release, fire and hazardous materials incidents (basicincident, fireincident, civiliancasualty, causes, codelookup). 2021. Accessed 2026-11-01.
• [4] U.S. Fire Administration (FEMA). NFIRS 2022 Public Data Release, fire and hazardous materials incidents (basicincident, fireincident, civiliancasualty, causes, codelookup). 2022. Accessed 2026-11-01.
• [5] U.S. Fire Administration (FEMA). NFIRS 2023 Public Data Release, fire and hazardous materials incidents (basicincident, fireincident, civiliancasualty, causes, codelookup). 2023. Accessed 2026-11-01.
• [6] U.S. Fire Administration (FEMA). NFIRS 2024 Public Data Release, fire and hazardous materials incidents (basicincident, fireincident, civiliancasualty, causes, codelookup). 2024. Accessed 2026-11-01.
• [7] U.S. Fire Administration (FEMA). NFIRS 2025 Public Data Release and About NFIRS & Release Notes 2025 (released 5/01/2026). 2025 (incomplete). Accessed 2026-11-01.
• [8] U.S. Fire Administration. NFIRS 5.0 Fire Data Analysis Guidelines and Issues. NFIRS 5.0. Accessed 2026-11-01.
• [9] U.S. Fire Administration. Cause for Structure Fires (Structure Fire Cause.pdf, PDR documentation). PDR 2024 documentation. Accessed 2026-11-01.
• [10] U.S. Fire Administration. NFIRS Data Snapshot: Thanksgiving Day Fires in Residential Buildings (2017-2019). 2017 to 2019. Accessed 2026-11-01.
• [11] U.S. Fire Administration. Topical Fire Research Series Vol. 3 Issue 2: Residential Structure Fires During the Winter Holiday Season. Published December 2002. Accessed 2026-11-01.
• [12] U.S. Fire Administration. Topical Fire Research Series Vol. 6 Issue 4: December and Holiday Fires. 2002 to 2004. Accessed 2026-11-01.
• [13] U.S. Consumer Product Safety Commission. 2023 Fireworks Annual Report (June 2024). 2023. Accessed 2026-11-01.
• [14] U.S. Consumer Product Safety Commission. CPSC Shares Fireworks Safety Tips as America Turns 250 (news release, June 24, 2026). 2025. Accessed 2026-11-01.
• [15] U.S. Fire Administration. National Fire Incident Reporting System (NFIRS) program page. Final NFIRS date Jan. 31, 2026. Accessed 2026-11-01.
• [16] U.S. Fire Administration. National Emergency Response Information System (NERIS) program page. 2024 to 2026. Accessed 2026-11-01.

Tier 3: journalism and expert commentary (1)

• [17] Denver7. Thanksgiving sees more than 3 times the daily average of home cooking fires (citing NFPA). Thanksgiving 2023. Accessed 2026-11-01.