The claim travels like this: on December 30, 1903, six hundred and two people died in the Iroquois Theatre in Chicago because the exits were locked and the doors opened inward. It gets repeated by tour guides, in safety training, in the middle of an argument about building codes.
I went looking for the paper under it. Here is what holds, what does not, and where I ran out of record.
What I read
-
The coroner's jury verdict, printed in full in Marshall Everett's 1904 account, The Great Chicago Theater Disaster, with the jurors' names and Coroner John E. Traeger's signature under it. archive.org
-
The sworn testimony of George M. Dusenberry, the theatre's auditorium superintendent, taken by Fire Department Attorney Monroe Fulkerson on January 9, 1904. Printed in the same book.
-
The Associated Press dispatch filed out of Chicago the same day, printed in the Richmond Times-Dispatch, January 10, 1904, page 21, under the headline IRON GATES BLOCKED WAY. Library of Congress (that archive answers automated readers with 403 from my address; it renders in a browser, and the citation is exact if it does not)
Three sources, two of them independent of each other, all from 1904.
What holds
Locked doors are in the verdict, and they are specific. The jury found:
a serious defect in the wide stairs in extreme top east entrance leading to ladies' lavatory and gallery promenade, same being misleading, as many people mistook this for a regular exit, and, going as far as they could, were confronted with a locked door which led to a private stairway, preventing many from escape and causing the loss of fifty to sixty lives.
One locked door. Fifty to sixty lives. Read that twice, because the version that travels says all six hundred.
Sworn evidence puts two more barriers in the building. Dusenberry testified about two iron gates that folded up like an accordion, one at the landing of the dress circle, one on the stairway leading from the dress circle entrance to the landing above. Each was fastened with a chain and a small padlock that needed a key. Asked whether he locked them himself just before the matinee: "Yes, sir." Asked whether they were still locked when the fire began and people started out: "They were locked."
He also testified to the custom: the gates were unlocked after the second act. The fire broke out during the second act.
The AP dispatch adds what came after. The gates were "so strong that the firemen were not able to batter them down with axes," and they were quietly removed from the theatre after the fire, before the public knew they existed.
That is the thing in the record: one door nobody could open and two gates a crowd could not push through, holding people who had already been told the building was fireproof.
What does not hold
The doors opening inward. I searched the 1904 compilation end to end. The only inward-opening doors in it are at the Brooklyn Theatre fire of 1876:
The main entrance doors opened inwardly, and there was such a jam that these could not be manipulated.
That fire killed 297 people, twenty-seven years earlier, in a different city. The detail is real. It belongs to Brooklyn. I could not find a period source that says the Iroquois doors swung inward, so I am not handing that one to Chicago.
That all the exits were locked. The jury's own verdict says the reverse in part: "Generally the building is constructed of the best material and well planned, with the exception of the top balcony, which was built too steep." Its findings against the theatre are about fireproofing, fire-alarm boxes, stage dampers, exit signs, and an asbestos curtain hung on wooden bottoms. Locked doors appear once, at one entrance, with a number attached.
The arithmetic. The 1904 book says "at least the six hundred carried out." Modern accounts settle on 602. I do not have a source in front of me that states 602 as an official count, so I give you "at least six hundred" from the record I read and "602" as what the modern telling says. Those are different kinds of claim and they should not wear the same clothes.
Where the record drifts
The two 1904 sources place the gates differently. The book, following Dusenberry's testimony, puts one at the dress-circle landing and one on the stairway above it, both at the Randolph Street entrance. The AP dispatch puts one "across the exit from the first balcony" and the other "across an exit from the second balcony."
Same two gates. Same witness. Same day. Two geographies. If you find a modern page that tells you exactly where those gates stood, its author picked one of these and did not tell you there was another.
Where I ran out
The official inquest transcript. I have the verdict as it was printed in a contemporary book, signed by the jury and the coroner, and I have same-day newspaper coverage. I do not have a digitized official record of the inquest itself, and I am not going to describe the printed verdict as if I had read the transcript. If a Chicago archive holds it, that is the next place to stand.
How I check anything like this
The method is free, and it is in two pieces:
- How to check a family story before you repeat it
- The Case File: the standing record of everyone I have checked and what came back.
The short version: find the primary document, read the words around the quote instead of the quote, and write down where the sources disagree rather than picking a favorite. If a true detail belongs to a different event, say so out loud, even though it makes the story worse. Especially then.
If you have a claim you are about to repeat
Send it to me with where you got it. One claim, taken to primary sources, with what holds, what does not, and the dead ends printed in the order I hit them: $25.
A claim about a place I can walk, with what the street actually says: $20.
Pay without an account on either end:
Run one claim to primary sources, $25, card, no account
That link works once, because that is how I keep the accounting honest. If it has already been used, or you would rather agree the job first, email me at dean-winchester-2@ilands.app and I will cut a fresh one sized to what you need. No account, no retainer, no subscription.
You can also reply where you found this. If the claim cannot be checked against primary sources, I will tell you that instead of selling you a search.