On April 19, 1995, a truck bomb brought down a large part of the north side of the Alfred P. Murrah Federal Building in downtown Oklahoma City. The bombing killed 168 people, and what that figure includes is set out below because the details matter. This guide leaves the crime to the courts and the historians. It covers how the response was organized: the local command and the flood of help in the first hours, the FEMA urban search and rescue task forces that came afterward, the risk of secondary collapse, and the access and credentialing problems that later federal doctrine tried to solve.

The Murrah Building at 9:02 a.m.: a transfer girder and a progressive collapse

The Murrah Building was a nine-story reinforced concrete federal office building on NW Fifth Street in downtown Oklahoma City, completed in the 1970s. Several federal agencies had offices there. Tenants included a Social Security office that drew members of the public, military recruiting offices and a day care center on the second floor. The bomb was in a rental truck parked at the curb on the north side, and it detonated at 9:02 a.m. local time on April 19, 1995. I had no part in this response. Everything here comes from the published record, chiefly three documents from 1996. The first is the City of Oklahoma City’s final report on the response. The second is the building performance assessment that FEMA and the American Society of Civil Engineers published. The third is the injury study the Oklahoma State Department of Health published in the Journal of the American Medical Association.

The structure explains much of what rescuers faced. On the north face, the columns of the upper floors did not run straight to the ground. They rested on a transfer girder at the third-floor level, which carried their loads onto fewer and more widely spaced columns at street level. The blast destroyed or badly damaged ground-level columns under that girder. Once the girder lost its support, the floors above came down. The building performance assessment described the result as a progressive collapse, in which the loss of a few critical members brought down far more of the building than the blast had damaged directly. The report also observed that the framing had little redundancy to carry loads around the failed members. That finding shaped later federal design guidance on progressive collapse.

The collapse left a debris field several stories deep against the north side, with floor slabs stacked and folded on top of each other. Above it stood the cut face of the surviving building, with partly attached floor sections hanging over the pile. As I read the health department’s study, the people in the collapsed portion were far more likely to die than the people in the parts of the building that stood. That fits the engineers’ view that the collapse, more than the blast pressure alone, decided who survived. It also means the rescue problem was mainly a structural collapse problem, worked underneath a damaged building that was still standing.

What the 168 counts: the children, the unborn and Rebecca Anderson

The Oklahoma City National Memorial and most official accounts give the death toll as 168 people. That number includes 19 children, many of whom were in the second-floor day care center. Most of the dead were inside the Murrah Building, though a small number were in neighboring buildings or outside when the bomb went off. The 168 figure is not seriously disputed. The disagreements come from what some sources add to it, so a writer citing the toll should say which of the following groups the number covers.

The first group is the unborn children of victims who were pregnant. Some family accounts and memorial materials name those children alongside their mothers. In the commonly published count they are not counted separately among the 168. A figure above 168 in a book, article or speech may include them, and the reader should check whether it does before assuming an error. I am not giving a number of pregnant victims here, because I have not confirmed one I trust. Anyone who needs that detail should take it from the memorial’s own records.

The second is Rebecca Anderson, a nurse who went to the scene to help. Contemporary news coverage and later memorial histories report that falling debris struck her on the head while she was assisting, and that she died several days later. Most published versions of the 168 do not include her, which is why some accounts give the toll as 168 people plus a nurse who died in the rescue effort. That is the account consistently reported, and I am repeating it as reported. For the official treatment of her name, consult the Oklahoma City National Memorial and Museum.

Injury totals are harder to pin down. Published figures run into the hundreds and vary because sources define an injury differently and count different populations, from people treated at hospitals to people who reported minor injuries later. I am not giving a single injury figure for that reason. The health department’s 1996 study is the most detailed epidemiological account, and anyone who needs a number should take it from that study along with its stated definitions.

The first hours: local command, a flood of help and the second-device evacuation

Oklahoma City Fire Department units arrived within minutes, and the fire department established and held incident command for the rescue. Police, the city’s ambulance service and the fire department were joined almost at once by a very large number of people who simply came. They included workers from damaged neighboring buildings, off-duty responders, responders from surrounding jurisdictions who self-dispatched, physicians and nurses from area hospitals, and passersby. The published accounts agree that the people removed alive were brought out by these local responders and bystanders on the first day, before any federal team could have arrived.

Medical care went into the debris along with the rescuers. The most widely reported case is that of Daina Bradley, a woman pinned in the rubble whose leg was amputated in place by a physician who went in to reach her, because there was no other way to free her. Hospitals meanwhile received many patients who arrived on their own or were brought by bystanders instead of going through field triage. That pattern has recurred in later mass casualty events, and it means the nearest hospitals see the first wave before EMS has a count.

During the morning, a report of a possible second explosive device led command to order everyone out of the building. Rescuers withdrew and left people still trapped while the area was checked. The object turned out not to be a live device. That evacuation is worth studying, because it showed both that command could clear the site when it had to and how much that order cost the people waiting in the debris. Withdrawal on a credible threat is still the correct call. The cost is the reason to have the order, the signal and the reporting process worked out ahead of time, so that the clearing takes as little time as possible.

The last survivor taken from the building is widely reported to have been a teenage girl, Brandy Liggons, removed on the evening of April 19. The published accounts agree that no one was found alive after the first night. That fact frames everything the federal task forces did in the weeks that followed.

The FEMA US&R system in 1995: local sponsors, a federal call-up and a standard roster

FEMA established its National Urban Search and Rescue Response System in 1989. The concept drew on experience with heavy rescue teams that some American fire departments, Fairfax County in Virginia and Metro-Dade in Florida among them, had built and deployed abroad. The system’s central feature is that the task forces are local. Each is sponsored by a state or local agency, usually a fire department, which recruits and trains the members and maintains much of the equipment. Under an agreement with FEMA, the sponsor deploys the team as a federal resource when FEMA activates it. In the Federal Response Plan of that era, urban search and rescue was one of the emergency support functions, with FEMA as the primary agency.

A task force is built to a standard roster. It includes search specialists and rescue specialists, canine search teams, medical staff including physicians and paramedics, and technical specialists. Those specialists cover structural engineering, hazardous materials, communications, logistics and planning. The team brings its own equipment cache and is meant to be self-sufficient for its first days on site, so it does not draw on a community that has just lost its infrastructure. The roster size and the typing of task forces have changed several times since the 1990s. Check FEMA’s current US&R program materials for today’s configuration instead of relying on any figure from that period.

The standard roster mattered at Oklahoma City because it let a local incident commander use teams from distant states without learning each one’s internal structure first. A task force from California and one from the East Coast arrived with the same positions, the same basic capabilities and leaders who knew what to ask for. That predictability eased the command integration problem. The system had deployed before 1995, including to the Northridge earthquake in 1994, but Oklahoma City put several of its task forces on a single urban building collapse together, under a local command, for an extended operation.

The task forces at the Murrah Building: rotation, search and the shift to recovery

The number of FEMA task forces most often cited for the Murrah response is eleven. They arrived from late on April 19 through the following days and rotated through the site, so that not every team was working at once. They worked under the fire department’s incident command rather than as a separate federal operation. Their job was to search the collapse systematically and remove debris and victims in a way that protected the people doing the work. Canine teams, search cameras and acoustic listening equipment let them check voids before committing rescuers. Debris removal combined hand work with heavy equipment, the balance depending on how close crews were to a possible victim and how unstable the surrounding structure was.

Because no one was found alive after the first night, the federal task forces spent their deployment on recovery. That changed the work in practical ways. Every victim recovered had to be documented and removed carefully, in coordination with the medical examiner, because identification depended on it. The building was also a federal crime scene, so recovery proceeded alongside evidence collection by the FBI and other federal investigators. Debris had to be handled with both purposes in mind, which slowed the work compared with a pure rescue operation. The city’s final report is the place to read how those three functions shared the same pile.

Search operations ended in early May. The remaining structure was brought down by controlled implosion on May 23, 1995. A small number of victims whose remains could not be reached safely while the damaged building stood were recovered after the implosion. That sequence is the clearest record of how seriously the structural hazard was taken, because command chose to leave remains in place rather than send crews under sections the engineers judged too dangerous.

Secondary collapse: the standing remainder, shoring and the authority to stop work

The secondary collapse hazard at the Murrah Building came from several directions at once. The surviving portion of the building stood directly over the debris pile with its damaged edge exposed, and partly attached floor sections hung from it. Loose concrete and building contents could fall from the damaged floors without warning. Removing debris from the pile could change how loads were carried within it, so taking out one slab might let another one shift. The reported injury that killed Rebecca Anderson, a blow from falling debris, is the kind of event the whole operation was organized to prevent.

The structural specialists on the task forces, working with other engineers at the site, assessed which areas could be entered and what had to happen first. Their main tools were shoring, built to carry loads that the damaged structure could no longer carry safely, and monitoring of the standing structure for movement. Standard US&R practice also posts lookouts with a defined evacuation signal, usually horn blasts, so that a crew deep in the pile knows to leave immediately without waiting for a radio call it may not hear. I am describing the doctrine here rather than claiming the specific signals used at the Murrah site, which the city’s report and the task force records would document.

The hardest part of managing secondary collapse is the authority to stop work. Somebody has to be able to order crews out on an engineer’s judgment, even while families are waiting and the crews themselves want to keep going. At Oklahoma City that authority ran through the incident command structure. The decision to leave some remains in place until after the implosion shows command accepting that cost. It ties back to the second-device evacuation on the first morning, because the same withdrawal mechanism serves both a structural warning and a security threat, and both depend on knowing who is inside so that command can confirm everyone came out.

Credentialing and access control: who belonged inside the perimeter

The convergence that saved lives in the first hours became a management problem within the first day. As I read the published accounts, a uniform, a set of turnout gear or hospital scrubs was effectively the credential during the early hours, and nobody could reliably say how many people were inside the perimeter or who they were. That matters for every hazard the site presented. A collapse warning or a security evacuation only works if command can account for everyone afterward, and accountability is impossible without knowing who went in. The site was also a crime scene and a place where human remains were being recovered, which raised the stakes of letting unknown people walk through it.

The accounts describe the perimeter being tightened and an identification and badging process established over the following days, so that access depended on an issued credential rather than an appearance. The accounts also describe unauthorized people getting inside before that system was in place. For the specifics of how badges were issued and checked, and which problems the city documented, see the city’s final report. Verifying volunteer clinicians posed its own problem, because an incident commander has no quick way to confirm that someone in scrubs is a licensed physician or nurse.

The federal task forces largely avoided this problem because they arrived as rostered units with known members, known qualifications and a leader accountable for each person. The credentialing gap at Oklahoma City lay with everyone else, including the self-dispatched mutual aid companies, the volunteer clinicians and the private equipment operators. That distinction is the useful lesson. A standardized, rostered resource is easy to admit and track. The hard part is a process for checking and badging people who were never part of any roster, and having that process exist before the day it is needed.

From Oklahoma City to later doctrine: what traces back and what does not

Some later federal changes trace fairly directly to the bombing. Presidential Decision Directive 39, signed in June 1995, set out federal counterterrorism policy. It separated crisis management, led by the FBI, from consequence management, led by FEMA, which was roughly the division of labor that had operated at the Murrah site between the criminal investigation and the response. Building security changed as well. The Department of Justice assessed the vulnerability of federal facilities in 1995, and an executive order later that year created the Interagency Security Committee to set security standards for federal buildings. Progressive collapse became a design concern in federal construction guidance, consistent with the building performance assessment’s findings.

Other connections are often claimed and should be stated more carefully. The Incident Command System did not come from Oklahoma City. It grew out of FIRESCOPE in California in the 1970s, and the fire department used it at the Murrah site because it already existed. The National Incident Management System, with its resource typing and its credentialing framework, came from Homeland Security Presidential Directive 5 in 2003, after the September 11 attacks. Oklahoma City is part of the experience behind NIMS, but describing NIMS as an Oklahoma City reform overstates the record. The US&R system itself continued and grew, and its task forces deployed to the World Trade Center and the Pentagon in 2001.

The fair reading is that Oklahoma City showed the national task force model working under a local command and exposed convergence and access control as unsolved problems. The World Trade Center response six years later showed the same convergence problem on a larger scale, which suggests the credentialing lesson had been recorded more than it had been absorbed. Current federal credentialing and resource typing guidance changes periodically. Verify current requirements with your state emergency management agency and FEMA before building anything on them.

What to do at your agency

• Have the special operations or technical rescue officer confirm in writing which FEMA US&R task force and which state or regional collapse rescue team your jurisdiction would request, and the exact request path through your state emergency management agency. That officer should add the result as one paragraph to the existing emergency operations plan this month.
• Have the training officer confirm that the collapse-site evacuation signal is written into the agency’s technical rescue guideline, and run one drill this month in which crews hear the signal, exit and report to a single accountability point within a timed limit.
• Have the emergency management coordinator write one paragraph into the existing emergency operations plan naming the person who issues site credentials at a major structural collapse, the identity and qualification proof that person accepts, and where the check-in point is set up.
• Have the fire chief add an agenda item to the next scheduled meeting with the county’s law enforcement leadership to name, by position, who serves as the law enforcement liaison to incident command when a collapse site is also a crime scene, and record the answer in the meeting minutes.
• Have the communications center supervisor confirm that the tactical channel plan for a structural collapse includes a channel or talkgroup reserved for site safety and evacuation orders, and that dispatchers have a written procedure for relaying an evacuation order to every unit on scene.
• Have the EMS medical director state in writing whether the agency has a process for getting a physician to a trapped patient inside a collapse zone, including who authorizes entry and who accompanies the physician, and send that statement to the special operations officer.

Takeaways

• The bombing of the Murrah Building on April 19, 1995 killed 168 people including 19 children, and the commonly published count does not include the unborn children of pregnant victims or Rebecca Anderson, the nurse reported killed by falling debris during the rescue.
• The 1996 FEMA and ASCE building performance assessment described a progressive collapse that began when ground-level columns under a third-floor transfer girder were lost, which made the response mainly a structural collapse operation.
• Local responders and bystanders removed the survivors on the first day, and the published accounts agree that no one was found alive after the first night.
• The FEMA US&R task forces, most often counted as eleven, worked under the Oklahoma City Fire Department’s incident command and spent their deployment on recovery alongside evidence collection and identification.
• Secondary collapse risk from the standing remainder of the building led command to leave some remains in place until after the controlled implosion on May 23, 1995.
• The rostered federal task forces were easy to admit and track, while self-dispatched responders and volunteers created the credentialing and accountability problems the response documented.
• Presidential Decision Directive 39 and federal building security standards trace fairly directly to the bombing, while ICS predates it and NIMS came from a 2003 directive after September 11.