CONELRAD was the first national plan in the United States for warning the public by broadcast radio. It began in 1951 as an air defense measure as much as a warning system, because its central purpose was to keep AM stations from serving as homing beacons for enemy bombers while still giving the public a place on the dial to find information. This piece explains how the plan was designed, what the documented record does and does not show about its use, and how it gave way to the Emergency Broadcast System in 1963 and the Emergency Alert System in 1997.

The 1951 problem: broadcast stations as bomber beacons

The Soviet Union tested its first atomic bomb in August 1949, and within a couple of years American air defense planning centered on the possibility of a manned bomber attack carrying nuclear weapons. Aircraft of the period routinely carried radio direction-finding equipment, and pilots had used AM broadcast stations as navigation aids for years because a broadcast transmitter sits at a known location, runs on a published frequency and stays on the air for long hours. A crew that tuned a direction finder to a station in a target city could take a bearing on it and fly toward it. Mediumwave signals also travel long distances at night by skywave, so a powerful station could be heard far from its home city.

The obvious defense would have been to shut every station off when an attack was detected, but that would also have cut off the public at the moment people most needed information. Broadcast radio was the one channel that reached nearly every household and many cars, and civil defense planning in those years assumed that the public would be told what to do by radio. The planners therefore needed an arrangement that would leave some broadcast service on the air for the population while making that service useless, or at least confusing, to a navigator in an enemy aircraft.

The arrangement was named CONELRAD, short for Control of Electromagnetic Radiation, and it was established in 1951 under the President’s authority over radio stations during war or national emergency, which the Communications Act of 1934 provides. The Federal Communications Commission administered the broadcast side of the plan in coordination with the military air defense organization, which was the source of the alert. Readers who want the exact executive and regulatory documents should go to the National Archives and the FCC’s historical records rather than rely on a citation from me, because I will not give document numbers I cannot confirm.

The on-air design: silence, two frequencies and clusters

The documented design worked in two parts. When an alert came, most broadcast stations, including FM and television stations, were to leave the air entirely. A smaller set of designated AM stations would stay on, but they would not stay on their normal frequencies. Each of them would shift to one of two frequencies in the AM band, 640 or 1240 kHz, and carry civil defense information instead of regular programming. A station that disappeared from its usual spot on the dial could no longer be identified by a navigator from a published frequency list, and a frequency shared by many transmitters across the country could not be tied to any single city.

The second part addressed what a direction finder would do with a shared frequency. Participating stations in an area were grouped into clusters, and the stations in a cluster took turns transmitting for short periods on the same frequency. A listener in the area would hear a continuous program, more or less, because the content was coordinated, while a direction finder on the aircraft would see the apparent source of the signal shift from one transmitter site to another. The intent was that no stable bearing would point to a target, so that homing on the CONELRAD signal would be impractical.

The design accepted a real loss of public coverage in exchange for that denial. Stations operating on a frequency other than the one their transmitter and antenna system were built around, and taking turns with other stations, would not reach their normal audiences, and listeners outside the reach of a participating cluster might hear nothing usable at all. Stations also had to be able to move to the new frequency quickly, which meant preparing equipment in advance, and the engineering details varied from station to station. I have not found a documented official rationale for choosing 640 and 1240 specifically that I would trust, beyond the fact that both sit inside the standard AM band and 1240 was already one of the channels used by many low-power local stations, so I will not offer one.

The listener’s part: triangles on the dial and an attention sequence

The public side of CONELRAD asked very little of the listener. A household was expected to own an AM radio, to know that in an emergency the information would be found at 640 or 1240, and to tune there when told. Radios sold in the United States during the CONELRAD years commonly carried small triangle markings at 640 and 1240 on the dial, using the civil defense symbol, so that a listener did not have to remember the numbers. Many of those radios survive in collections and antique shops, and the triangles are the most visible physical trace the program left. I have seen the markings described as legally required and also as standard industry practice, and I have not confirmed which description is correct, so I describe them only as common.

Getting stations to act together required an alerting method that reached broadcast engineers before it reached the public. Period descriptions of CONELRAD describe a prescribed sequence in which a station interrupted its carrier in a set pattern and then transmitted a steady tone, and other stations that monitored it were expected to recognize the sequence and carry out their own CONELRAD procedures. I am not going to reproduce the exact timings from memory, because I have seen them stated slightly differently in different sources. The principle matters more than the timings. Stations listened to designated stations, and the alert moved through the broadcast industry by that monitoring, rather than over a dedicated network to every transmitter.

That monitoring arrangement put the reliability of the whole system in the hands of the person on duty at each station, who had to be listening, recognize the signal and follow the procedure. It also assumed that the public would have a working AM receiver within reach of a participating station, with batteries if the power was out. None of those assumptions was unusual for 1951, but each of them reappears in later systems, and the monitoring relay in particular survived into both EBS and EAS in modified form.

The record of use: drills, tests and an untested premise

CONELRAD was exercised in tests and in the national civil defense drills of the 1950s, and broadcast stations had to be prepared to carry out its procedures. That is the documented extent of its operational life. It was never used for an actual attack, because there was no attack. A repeated notion holds that CONELRAD was widely used in practice, that families regularly heard it come on during crises, or that it carried warnings through the confrontations of 1960 to 1962, and that notion is a myth as far as the record I can find supports. I have seen retellings that place a CONELRAD activation in one specific Cold War crisis or another, and I cannot find a documented national activation for any of them, so I am not repeating those stories here.

The 1960 period is where the retellings tend to cluster, probably because the U-2 incident, the Berlin tensions and the Cuban Missile Crisis gave the era a sense of constant emergency that later writers project back onto the warning system. Heightened readiness at military commands, which is documented for some of those episodes, is a different thing from a CONELRAD activation that took stations off the air and sent them to 640 and 1240. Readers who come across a claim that the system was activated on a specific date should ask for the FCC record or the contemporary press account before accepting it, and should treat a memoir or a museum caption on its own as a recollection rather than a finding.

Because the system was never activated for its purpose, two questions remain open. The first is whether the frequency shift and clustering would have frustrated an enemy navigator in a real raid. The second is whether the public would have received useful information on two shared frequencies with reduced coverage. Both are questions about an untested design, and I have not found an authoritative evaluation that settles either one. What the record does settle is the design itself, the dates of the program, and the reason it ended, which turned out to have nothing to do with how well clustering worked.

1963: missiles end the beacon argument and EBS takes over

CONELRAD’s whole technical premise was that the threat was an aircraft with a navigator who needed a radio reference to find a target. By the late 1950s and early 1960s the strategic threat was moving toward intercontinental ballistic missiles, which did not navigate by homing on broadcast stations. Once missiles became the primary concern, the benefit of moving stations to two shared frequencies largely disappeared, while the costs to public coverage remained. Federal planners took the logical step and dropped beacon denial as the organizing purpose of broadcast warning.

The Emergency Broadcast System replaced CONELRAD in 1963. The most important change for the public was that participating stations no longer had to leave their normal frequencies. A listener could stay tuned to the station already on the radio and expect emergency information there, and the reasoning behind the 640 and 1240 markings no longer applied. The national purpose stayed in place: EBS existed so that the President could address the country by broadcast during a national emergency, with stations relaying the message down through the monitoring structure.

Several features carried over from CONELRAD. Stations still monitored designated stations, an attention signal still preceded emergency content, and the system still depended on staff at each station recognizing an activation and acting on it. EBS also refined the attention signal over time, and the two-tone signal that many readers will remember, combining tones at 853 and 960 Hz, came into use during the 1970s. That sound was designed to be distinctive enough to get attention and to be detected by station monitoring receivers, and it remained the public’s audible marker of EBS until the system was retired.

EBS in service: weekly tests, local weather and the 1971 false activation

For most Americans who grew up between the 1960s and the 1990s, EBS was something heard during tests. Broadcast stations ran regular tests in which an announcer explained that the station was conducting a test of the Emergency Broadcast System, sent the attention signal, and explained that in a real emergency the listener would have been told where to tune and what to do. Over time EBS also came to be used, on a voluntary basis, for state and local emergencies, most often severe weather, and that local use was the version of the system most people would actually have heard outside a test.

The national activation path, the system’s reason for existing, was never used for an actual national emergency. It was, however, triggered by mistake. In February 1971 an operator at the NORAD facility that originated national EBS messages sent an actual emergency activation message instead of the scheduled test, and the cancellation took a considerable time to go out. Many stations did not act on the message, and the FCC and the press at the time examined why. I am leaving out the precise number of minutes and the operational details because accounts differ on them; the FCC’s records from 1971 and contemporary press reporting are where to read the specifics.

The 1971 episode and routine experience pointed to the same weaknesses, and those weaknesses shaped what replaced EBS. Activation depended on a person at every station hearing the signal, judging whether it was real and relaying it by hand, which introduced delay and inconsistency at each link in the chain. The attention signal carried no information about the event, the location or the duration, so a station receiving it could not tell automatically whether the message applied to its own audience. Stations that were unattended, which became more common as automation spread through broadcasting, were poorly matched to a design that assumed a human was always at the console.

1997: EAS, digital headers and unattended relay

The FCC adopted rules for the Emergency Alert System in the mid-1990s, and broadcast stations were required to have EAS equipment in place by January 1, 1997, which is the usual date given for the end of EBS. Cable systems were brought into EAS on a phased schedule afterward. The central technical change was that an alert now began with a short digital header instead of a bare attention tone. The header used the Specific Area Message Encoding format that the National Weather Service had developed for NOAA Weather Radio, and it identifies the originator of the message, the type of event, the geographic areas affected, how long the alert is valid and the station relaying it.

Because the header carries that information in machine-readable form, EAS equipment at a station can decide for itself whether an alert applies to its coverage area and can forward it automatically, without waiting for a human to recognize a tone and act. Each station monitors more than one source under its state’s EAS plan, so a failure at one upstream station does not silence everyone below it. State Emergency Communications Committees write those plans and assign the monitoring relationships, and the plans are the place to look for how alerts actually move in a particular state. The national path still exists, carried through a set of broadcast stations FEMA designates as Primary Entry Point stations, which are hardened and equipped with backup power so that a presidential message can reach the rest of the system.

EAS has continued to evolve since 1997, most notably through FEMA’s Integrated Public Alert and Warning System, which lets authorized originators send alerts in the Common Alerting Protocol format to EAS and to other channels. The first nationwide EAS test took place on November 9, 2011, and FEMA and the FCC identified a range of problems in it that later tests were meant to address. The FCC’s published reports on the nationwide tests are the authoritative record of what went wrong and what changed, and anyone responsible for local alerting should read them rather than rely on summaries.

What carried forward: the assumptions an agency still inherits

The most durable feature of CONELRAD is the monitoring relay, in which broadcast stations listen to designated sources and pass an alert onward. CONELRAD did it with carrier interruptions and a human ear, EBS with a two-tone signal and a human hand, and EAS with a digital header and equipment that can relay without a person present. The chain is still there, written into state EAS plans, and a local alert that is meant to reach broadcast audiences depends on the chain being configured correctly between the originator and the stations that serve the agency’s jurisdiction.

The second inheritance is the original national purpose, which was a presidential message during an attack and has never been used for that purpose across three generations of systems. Meanwhile the local and weather uses, which began as an add-on, became the work these systems actually do. That pattern matters to a local agency because the parts of the system most exercised are not always the parts built for its own needs, and the procedure a county uses to request or originate a local alert is often less practiced than the national test schedule.

The third inheritance is the assumption that the listener has a receiver turned on and tuned. CONELRAD assumed an AM radio and a household that knew where to tune, and broadcast-based EAS still reaches only people listening to or watching a participating outlet, which is why wireless alerts to phones and NOAA Weather Radio now carry so much of the load for people who are not. An agency that understands the history will treat the broadcast path as one channel with known limits, and will confirm through the state EAS plan and its own IPAWS arrangements that a local alert actually reaches the stations its residents listen to.

What to do at your agency

• The emergency management director requests the current state EAS plan from the State Emergency Communications Committee chair or the state emergency management agency this month and files it with the agency’s warning annex, noting the Local Primary stations listed for the operational area that includes your jurisdiction.
• The emergency management coordinator reads the local alert origination section of the state EAS plan and writes one paragraph into the existing warning annex stating exactly who at your agency can originate or request a local EAS alert and by what path.
• The person who holds the agency’s IPAWS alerting authority confirms this month that the memorandum of agreement with FEMA is current and that every named originator has completed the required training, and records the expiration dates in the annex.
• The communications center supervisor adds an agenda item to the next scheduled meeting with the local broadcast station engineers asking them to confirm which sources their EAS equipment monitors and whether your jurisdiction’s location codes are configured to relay.
• The radio system administrator checks that the emergency operations center has a working battery-powered AM and FM receiver, tunes it to the Local Primary stations named in the state plan, and logs that both can be received inside the building.
• The training officer adds a fifteen-minute review of the local alert request procedure, including the backup path if the primary originator is unavailable, to the next scheduled telecommunicator in-service.

Takeaways

• CONELRAD began in 1951 as a plan to keep AM stations from serving as homing beacons for enemy bombers while still giving the public a broadcast source of emergency information.
• Under the documented design, most stations left the air during an alert while designated AM stations shifted to 640 or 1240 kHz and took turns transmitting in clusters to frustrate airborne direction finding.
• The notion that CONELRAD was widely used in practice, or activated during specific crises of 1960 to 1962, is a myth as far as the documented record shows, because the system was exercised in tests and drills but never activated for an actual attack.
• Whether clustering would have defeated direction finding in a real raid was never tested, so claims about the system’s effectiveness describe design intent rather than an outcome.
• The Emergency Broadcast System replaced CONELRAD in 1963 because ballistic missiles removed the beacon-denial rationale, and stations were then able to carry emergency information on their own frequencies.
• The Emergency Alert System replaced EBS with an equipment deadline of January 1, 1997, adding a digital header that identifies the event, the area and the originator and allows automatic relay.
• The monitoring relay that began with CONELRAD still governs how alerts reach broadcasters, and the state EAS plan is the document that shows a local agency how its own alerts travel.