On the evening of October 8, 1871, fire swept through Peshtigo, Wisconsin and the farm settlements around it, killing more people than any other fire in American history. The same night, and driven by the same weather, Chicago burned. Chicago got the correspondents, the relief money, the building codes and eventually a national observance named after it, while the deadlier fire 250 miles north has never had a settled death toll. This piece covers what is documented about that night, why the published toll ranges from roughly 1,200 to 2,400, how a mass fire behaves, and what the attention gap means for a county emergency manager today.

What burned on October 8, 1871

Peshtigo sat on the Peshtigo River in Marinette County, a few miles up from the waters of Green Bay, and it was a company town built around lumber. The Peshtigo Company operated a sawmill and a large woodenware factory there, and because the village was built out of what the mill produced, the sidewalks and the buildings were wood and the streets were surfaced with sawdust. The population of the village is usually given at somewhere between 1,700 and 2,000 people. Fires had been burning in the surrounding cutover land for weeks, which was normal enough that nobody treated it as an emergency. The fire that destroyed the village arrived from the southwest in the late evening of Sunday, October 8, and accounts generally place its arrival between about nine and ten o’clock, with the village substantially gone inside an hour or two.

The village was not the whole event. The farm settlements southwest of town known as the Sugar Bush burned, Menekaunee and parts of Marinette burned, and fires ran at the same time on the east side of Green Bay through Door and Kewaunee counties, where the settlement of Williamsonville was destroyed at a site that Door County later preserved as Tornado Memorial Park. The total area burned in northeastern Wisconsin and across the bay is commonly published at roughly 1.2 million acres, about 2,400 square miles, and some accounts go higher. Across Lake Michigan the same night, fires hit Holland, Manistee, Port Huron and the Thumb region of Michigan, in events usually grouped together as the Great Michigan Fire, with deaths often estimated at around 200 and counted no more carefully than Wisconsin’s were.

The Chicago fire began that same evening and burned into October 10, destroying roughly 2,100 acres and something on the order of 17,000 to 18,000 buildings, leaving about 100,000 people homeless, and killing a number usually given as about 300, which is itself an estimate rather than a count. One man held property in both places, because William B. Ogden, the first mayor of Chicago, was also president of the Peshtigo Company, and he lost heavily in each fire on the same night.

Drought, slash, and the weather system that arrived that evening

The summer and fall of 1871 were extraordinarily dry across the upper Midwest. Contemporary accounts describe wells and streams going down, swamps drying out so that peat and duff could carry fire underground, and almost no measurable rain over the region from midsummer into October. Dryness alone does not produce a fire that kills a thousand people, because the fuel has to be there in quantity and arranged so that it burns all at once, which is exactly what the logging economy of the period had created.

Clear-cut logging in that era took the merchantable trunk and left the tops, limbs and bark on the ground, so the cutover land was covered in cured slash several feet deep over large areas. Railroad construction crews extending line through the region burned slash to clear right of way, farmers burned to clear land for planting, and the mills and the town itself were built out of the same material. Small fires burned continuously through late September and early October, and residents of Peshtigo had been smelling smoke for weeks, which is the single most important detail for understanding why nobody evacuated. The hazard had been present and tolerable for so long that its change in character was not recognized until the fire was on the edge of town.

What changed on October 8 was the weather. A strong low pressure system moving east across the Plains and the upper Midwest tightened the pressure gradient over Wisconsin, Illinois and Michigan, producing strong, gusty winds out of the southwest and a cold front crossing the region that evening. Those winds are the common factor linking the Chicago fire, the Peshtigo fire and the Michigan fires, which were not one fire and were not coincidences but separate ignitions in separate fuel beds that all blew up under the same synoptic setup within hours of each other. Rain over the region on October 9 and 10 is generally credited with helping to end all of them.

What a mass fire does that a brush fire does not

A fire large enough and hot enough stops being a line of flame pushed along by the ambient wind and starts generating its own circulation. The buoyant plume over a very large burning area rises hard, and the air that replaces it has to come in across the surface, which produces indraft winds at ground level that can exceed anything the weather map would support. Fire researchers generally distinguish a firestorm, with a roughly stationary convection column drawing air inward from all sides, from a conflagration with a moving front driven by the prevailing wind, and the accounts from Peshtigo describe features of both: a wind-driven run into the town, and then localized behavior in which the fire made its own weather over the village.

Several consequences follow that matter operationally. Rotating columns form when horizontal vorticity near the surface is tilted into the vertical by the rising plume, and the resulting fire whirls can lift burning debris, roofing and whole sections of wall, which is why survivor accounts from Peshtigo describe buildings coming apart upward and why the Door County memorial park carries the word tornado in its name. Radiant and convective heating preheat fuels well ahead of the flame, so ignition does not wait for contact, and spotting puts new fires hundreds of yards to miles downwind. Fires burned on the Door Peninsula across the waters of Green Bay the same night, and accounts differ over how much of that was burning material carried across and how much was separate fires already in the ground there, which is a distinction nobody can resolve now.

The other consequence is that the atmosphere near the ground becomes lethal independent of flame. Superheated gases, high carbon monoxide and depleted oxygen kill people who never touch fire, and that mechanism accounts for deaths in open fields and on roads at Peshtigo where the bodies were not badly burned. Peshtigo has been studied as a reference case for mass fire behavior, including in Cold War era civil defense research comparing historical conflagrations with the fire effects expected from nuclear weapons, alongside Hamburg in 1943 and Tokyo in 1945. The same cutover-land conditions kept producing the same result for decades afterward, with the Hinckley, Minnesota fire of 1894 and the Cloquet fire of 1918 carrying commonly published death tolls of 418 and 453.

The death toll, and why it cannot be settled

There is no authoritative number for how many people died at Peshtigo and in the surrounding country that night, and anyone who hands you one without a range is giving you a number that somebody picked. The range most often published runs from about 1,200 at the low end to about 2,400 at the high end, with Wisconsin state historical sources and National Weather Service material on the fire both describing the toll in roughly those terms, and some popular accounts going to 2,500. The low figure generally traces back to bodies recovered and known missing persons compiled in the weeks afterward by relief officials and local clergy. The high figure comes from estimates of how many people were in the burned district in total, measured against how many came out of it.

The recordkeeping failed for reasons that are specific and worth naming. The town’s records burned with the town. A large share of the population consisted of recent immigrants, many of them Belgian, German, Scandinavian and Irish, who had arrived since the 1870 census and were on no local roll. Lumber camps and railroad construction crews in the woods held men whose employers did not keep usable rosters and in some cases did not survive to be asked. Entire family groups in the Sugar Bush settlements died together, leaving nobody to report them missing, and the state of Wisconsin had no agency in 1871 charged with producing a casualty count. Roughly 350 unidentified victims were buried in a mass grave at the Peshtigo Fire Cemetery, which is the physical expression of the counting problem.

Whatever the true figure, the comparison that matters holds under any version of it. The Chicago fire’s commonly cited toll of about 300 is a fraction of even the lowest Peshtigo estimate, and the Peshtigo fire remains the deadliest fire of any kind in United States history by a wide margin, ahead of Hinckley, Cloquet and every structure fire in the record. When you write the number into a slide or a plan, write the range, say who publishes it, and say plainly that the records to settle it were destroyed in the event itself.

How people died, and the stories that grew afterward

The Peshtigo River is the reason there were any survivors in the village at all. People went into the water and stayed there for hours, dousing their heads, pushing away burning debris and floating logs, with livestock in the water alongside them. Deaths in the river came from drowning, from exhaustion, and afterward from exposure, because the night turned cold on people who were soaked and had lost their homes and their clothing. Others died in root cellars and in wells, where the fire consumed the oxygen above them, and a number of people who ran for cleared fields were found dead in the open without severe burns, which is consistent with superheated air and carbon monoxide rather than flame contact.

Father Peter Pernin, the Catholic priest serving Peshtigo and Marinette, survived in the river and later published a firsthand narrative of the night that has been republished by the Wisconsin Historical Society Press as The Great Peshtigo Fire: An Eyewitness Account. It is the single most useful primary source on what the fire looked like from inside the town, and anyone building a training presentation on this event should read it directly rather than rely on secondary summaries, including this one.

Two persistent claims about Peshtigo should be labeled as myths in the same breath as they are repeated. The first is the idea that the fires of October 8 were ignited by fragments of a comet, a claim advanced by the Minnesota writer and former congressman Ignatius Donnelly in his 1883 book Ragnarok: The Age of Fire and Gravel and recycled periodically since. It does not hold up, because meteoritic material reaching the ground is not a reliable ignition source, and every one of the fires that night had ordinary ignition already burning in the fuel bed for weeks beforehand. The second is the tidier claim that Peshtigo was simply forgotten because nobody cared, which understates a real structural difference in how news, money and institutional memory moved in 1871.

Why Chicago got the attention and Peshtigo did not

Chicago was a national financial and rail center with resident correspondents for eastern newspapers, telegraph connections that were restored within days, and a property loss that was documented in detail because much of it was insured. Peshtigo had one telegraph line into the region, it burned, and the first reliable word of what had happened traveled by boat and messenger to Marinette and Green Bay and then onward, by which time the Chicago story had been on the wires for a day or more and had claimed the available attention. Wisconsin accounts credit Frances Fairchild, wife of Governor Lucius Fairchild, with redirecting a relief shipment northward while the governor was in Chicago attending to the relief effort there, which captures the proportions of the moment accurately.

Money followed the same channel as the news. Chicago received donations from across the United States and from Europe, and the fire’s effect on the insurance industry, which it bankrupted in large numbers, drove decades of work on construction standards, water supply and fire underwriting. The Peshtigo losses were mostly uninsured, belonging to immigrant laborers and farmers whose assets were a house, a team and a few cleared acres, so they generated no claims data, no actuarial lesson and no sustained institutional interest. The National Fire Protection Association, which observes Fire Prevention Week in the week containing October 9 and traces the observance to Fire Prevention Day proclaimed by President Woodrow Wilson in 1920 and to the first National Fire Prevention Week proclaimed by President Calvin Coolidge in 1925, named the observance for the Chicago anniversary.

The mechanism generalizes, and it is the part worth carrying into a modern EOC. Attention is allocated by who can report, who can be reached, who has an institutional constituency that keeps raising the subject afterward, and whose losses were measured in a form that the record keeps. A rural county with a destroyed communications path, a population that does not generate claims data, and no full-time public information officer will be underrepresented in the national account of an event no matter how severe its losses were, and underrepresentation in the account turns into underrepresentation in the relief and in the mitigation funding that follows.

Two incidents, one pool of resources

The night of October 8, 1871 is the earliest large American example of a problem that every state emergency management agency still works: two or more significant incidents running at the same time, driven by the same weather, competing for one pool of resources and one set of decision makers. The weather pattern that creates your fire danger also creates your neighbor’s, which means a red flag warning or a critical fire weather outlook covering several hundred miles can commit every regional partner you have on the same afternoon, and the same logic applies to flooding along a shared basin and to wind events crossing several states.

The wildland fire system in the United States has a formal answer to this, which all-hazards agencies often lack. Geographic Area Coordination Centers track committed resources and set regional priorities, the National Multi-Agency Coordinating Group at the National Interagency Fire Center sets national preparedness levels and prioritizes incidents when demand exceeds supply, and at the higher preparedness levels the allocation of crews, aircraft and overhead is an explicit national decision rather than a first-come arrangement. Agencies working an incident under that system should understand that when they are told no resources are available, a documented prioritization decision stands behind that answer, and they can ask where their incident sits in it.

Outside the wildland fire system, the equivalent decision usually belongs to a state EOC operations or logistics section, sometimes with a multi-agency coordination group convened for the purpose, and the process is frequently less visible and less documented. Most emergency managers I have worked with cannot name, from memory, who in their state makes the call when two counties request the same scarce asset in the same hour, and that is a question with a real answer that can be obtained in a phone call before the day it matters. Contemporary cases make the point without any need to speculate about motive, since the Camp Fire and the Woolsey Fire both ignited on November 8, 2018 in California and ran simultaneously, with Cal Fire publishing 85 civilian deaths for the Camp Fire and three for Woolsey, and both drew on the same state and federal resource pool.

Knowing who was there before you have to count them

The reason Peshtigo’s toll is a range rather than a number is that nobody knew the denominator. The village had a population estimate, the lumber camps and railroad crews in the woods did not, the Sugar Bush farms held households that had arrived since the last census, and the documents that might have resolved it burned. Every modern county has some version of the same gap, because the resident population on the census is not the population present on a given night.

The categories that produce the gap are identifiable in advance. Seasonal agricultural workers in labor housing, construction and pipeline crews in temporary camps, campgrounds and RV parks in season, short-term rentals that no agency licenses in many jurisdictions, university housing that is full or empty depending on the calendar, correctional and detention populations, long-term care residents, and people living unsheltered all move the real nighttime population away from the planning figure. The Camp Fire demonstrated how hard the resulting accounting is even with modern tools, because the Butte County Sheriff’s Office published a list of unaccounted-for people in November 2018 that at one point carried more than a thousand names and was revised repeatedly as individuals were located, which is the normal behavior of such a list and not evidence that anybody did the work badly.

The practical work is unglamorous and can be done in an afternoon. Assessor parcel data and utility customer counts give a structure-level denominator for an evacuation zone. The licensing or parks office can produce a list of campgrounds and labor camps with capacity. The school district knows its enrollment by building, and the long-term care facilities know their census daily. Getting those into the evacuation and mass care annexes before an event, with a date on each, is what makes it possible to say later how many people were in a zone, and the absence of that work is what leaves historians arguing over a range 150 years afterward.

What to do at your agency

• Have the emergency manager obtain, in writing, the name and title of the person or group in your state who prioritizes competing resource requests from multiple jurisdictions in the same event, and add that entry to the EOC call list this month.
• Ask your county GIS or assessor’s office for a parcel and structure count inside each named evacuation zone, attach it to the evacuation annex with the date it was produced, and set a calendar reminder to refresh it annually.
• Have the fire chief or operations chief pull the current mutual aid roster, mark every partner that sits inside the same National Weather Service fire weather zone or the same red flag warning area as you, and start one new agreement with a department outside that area before the end of the quarter.
• Have the communications supervisor test one alternate reporting path to the state EOC that does not use commercial telephone or internet, whether that is a statewide radio talkgroup, a satellite phone or an amateur radio liaison through your local ARES group, and log the date and the result of the check-in.
• Ask your licensing office, parks department or code official for a current list of campgrounds, RV parks, seasonal labor housing and short-term rental properties with capacities, and file it with the mass care annex.
• Write one paragraph into the existing public information annex stating that all early casualty and damage figures released by your agency carry the words preliminary and a timestamp, and naming the role responsible for issuing the correction when the figure changes.
• Put a fifteen-minute item on an officers’ meeting that already happens: the scenario is a wind-driven fire in your interface area on the same afternoon that the rest of the region is committed, and the crew states out loud what they do with the apparatus they have and who they notify first.

Takeaways

• The fire that destroyed Peshtigo, Wisconsin on the evening of October 8, 1871 remains the deadliest fire in United States history, with published death toll estimates ranging from roughly 1,200 to 2,400 and some accounts going to 2,500.
• The range cannot be narrowed because the town’s records burned, much of the population had arrived since the 1870 census, and lumber camps and railroad crews in the surrounding woods were on no usable roster.
• The Great Chicago Fire burned the same night, killed a number usually given as about 300, and covered roughly 2,100 acres against something on the order of 1.2 million acres in northeastern Wisconsin and across Green Bay.
• Both fires, along with the Michigan fires at Holland, Manistee and the Thumb region, were separate ignitions that blew up under the same strong southwesterly winds from a low pressure system crossing the region on October 8.
• A mass fire generates its own indraft winds, rotating columns and long-range spotting, and it kills through superheated gases and oxygen depletion well away from direct flame contact.
• The claim that comet fragments ignited the fires, advanced by Ignatius Donnelly in 1883, is a myth, because every one of those fires had ordinary ignition sources already burning in the fuel bed for weeks.
• Chicago’s attention advantage came from working telegraph lines, resident correspondents, insured losses that generated documentation, and an industry that kept studying the event, none of which Peshtigo had.
• Simultaneous incidents competing for one resource pool is a current operational problem, and the wildland fire system answers it through Geographic Area Coordination Centers and the National Multi-Agency Coordinating Group while all-hazards agencies often have no documented tie-breaker at all.