Somewhere over Wyoming in the fall of 1927, a mail pilot throttled back and looked down at ten thousand square miles of nothing. No highway cut through the basin below him. No radio crackled anything useful into his headset, because there was no radio worth trusting yet. There was a compass, a wristwatch, a chart folded into his flight jacket, and the quiet fear that came from knowing that if his engine quit here, nobody would find him for a long time.

Then he saw it: a pale yellow shape on the desert floor, long as a tennis court, unmistakably geometric against the scrub. It isn’t an abandoned runway, and it has nothing to do with aliens. These giant concrete arrows were part of America’s Transcontinental Airway System, the federal navigation network that guided airmail pilots coast to coast decades before GPS.

It is a signpost. They were meant to be so obvious, so impossible to miss from a thousand feet up, that the federal government built one of the largest pieces of hidden infrastructure ever poured onto the American landscape: a visual highway in the sky when you have no lights, no radio, and no way to see the ground at night, using nothing but concrete, paint, and a very simple idea.

If you’ve ever seen photographs of these concrete arrows America built, you already know the basic shock of them. They show up on hiking forums, in Google Earth listicles, in off-road Jeep club trip reports from Nevada and Utah. They’re strange enough that people have proposed, half-jokingly, that they’re relics of some vanished civilization, or a forgotten segment of Route 66. None of that is true, but the confusion is understandable, because the real explanation requires you to imagine an America that doesn’t quite match the one in your head: an America where the government spent the 1920s physically painting a route across the continent so pilots could see, with their own eyes, where to fly next.

These arrows are the surviving skeleton of the Transcontinental Airway System, and they were never meant to be mysterious. They were meant to be so obvious, so impossible to miss from a thousand feet up, that a nervous pilot with a few years of flying experience could find his way from New York to San Francisco without getting lost over Nebraska.

What’s actually strange is not that the arrows exist. It’s that almost nobody remembers why.

To understand why the government built a two-thousand mile chain of concrete signposts, you have to go back to a period when airmail was less a service than an extended national argument about whether flying the mail was worth the body count.

The U.S. Post Office Department began regular airmail service in 1918. It was, from the start, a hybrid operation, because nobody had solved night flying yet. Planes flew during daylight and handed the mail off to trains after dark, which meant the promised speed advantage over rail evaporated the moment the sun went down. Thirty-four of the approximately 200 pilots who flew the mail between 1918 and 1927 lost their lives in crashes, a staggering fatality rate for a civilian workforce. Pilots and postal workers grimly nicknamed it the “Suicide Club.”

The fundamental problem was navigation, not aircraft technology. Airplanes of the era could physically fly at night without much modification. What they couldn’t do was find their way. There was no reliable radio navigation. There were no aeronautical charts worth the name: pilots often flew with automobile road maps folded on their laps, looking for towns, rivers, and rail lines as landmarks. Dead reckoning, essentially educated guessing based on compass heading, airspeed, and elapsed time, was the only tool available for the stretches where the ground gave you nothing to look at. Get the wind wrong by a few degrees over a four-hour leg and you could miss an entire state.

The pressure to solve this wasn’t only operational. It was political. In February 1921, Postmaster General Albert Burleson staged a demonstration flight explicitly designed to save the airmail program from a newly elected Congress that was talking openly about shutting it down. Relay teams launched from both coasts, flying day and night, to prove continuous air delivery was faster than rail. It nearly fell apart. One pilot, William Lewis, died in a crash near Elko, Nevada, during the attempt. By the time the westbound relay reached North Platte, Nebraska, weather had grounded almost everyone else.

That’s where a pilot named Jack Knight enters the story, not as a footnote, but as the person whose single flight arguably created the political will for everything that followed. Much like railroads standardized distance, Knight’s flight forced a chaotic layout into a rigorous national path.

Knight had already flown one scheduled leg that night, landing at Omaha with a fresh blizzard rolling in and a broken nose from a crash the week before. His relief pilot took one look at the weather and refused to fly. Rather than let the relay die there, Knight climbed back into his de Havilland and kept going, through country he had never flown, in the middle of a snowstorm, using a paper map somebody had literally torn off the office wall and handed him on his way out the door.

He navigated by watching for the glow of bonfires that postal workers, farmers, and airfield staff had lit along his route in Nebraska and Iowa, town by town, without knowing whether he’d actually make it that far. “I felt as if I had a thousand friends on the ground,” he said afterward, recalling the string of small-town fires sliding past beneath his wings. He landed in Chicago at dawn to a waiting crowd, having flown roughly 830 miles through darkness, snow, and unfamiliar terrain on nothing but nerve and a torn map. The Omaha Bee reported the next morning that Knight was “the man who flew with the mails from Cheyenne to Chicago in a single night,” cementing his status in aviation history. The full relay delivered mail from San Francisco to New York in 33 hours and 20 minutes. Congress, suitably impressed, kept the funding flowing.

Bonfires lit by volunteers were never going to scale into a national system. They depended on goodwill, good timing, and good luck, three things no infrastructure planner wants to build a program around. What the country needed was something permanent, professional, and indifferent to whether a farmer remembered to light his fire that night. Knight himself later worked with postal officials and civic leaders to help establish the first generation of permanent navigational beacons, the direct ancestor of the system that would eventually crisscross the continent.

Congress responded with the Air Mail Act of 1925 and the Air Commerce Act of 1926, and in 1926 to 1927 the Department of Commerce formally took over responsibility for the nation’s airways from the Post Office. What they built next is the actual subject of this article.

The system that emerged, formally called the Transcontinental Airway System, was proposed by the National Advisory Committee for Aeronautics, funded by Congress, and built and managed by the Department of Commerce’s Aeronautics Branch. The first segment went up between Chicago and Cheyenne, Wyoming, deliberately positioned in the middle of the route so aircraft leaving either coast in daylight could reach the lighted airway by nightfall. This initial deployment anchors the entire route, acting like a structural baseline for early American infrastructure. Night service on that stretch began July 1, 1924, cutting transcontinental delivery time by two full days by eliminating the nighttime rail handoff entirely.

The engineering was not glamorous, but it was thorough. At intervals of roughly 10 to 15 miles, closer together in mountainous terrain, farther apart on open plains, crews erected steel towers, usually around 51 to 53 feet tall, built from prefabricated angle-iron sections bolted together on site. Atop each tower sat a rotating beacon rated at roughly 5 million candlepower, visible under good conditions for up to 40 miles. Fixed “course lights” pointed toward the neighboring stations on either side, so a pilot could tell “here’s an airport, keep going” from “this is just a waypoint, don’t land.”

Each tower stood at the center of a poured concrete arrow, typically 50 to 70 feet long, painted bright yellow for daytime visibility. The arrow pointed toward the beacon carrying the next-highest number along the route, so a pilot flying a given Contract Air Mail route simply had to keep tracking arrows in ascending order to stay on course. Where no electrical grid existed to power the light, which described most of the rural West in the 1920s, the Department of Commerce installed a generator inside a small shed at the base of the tower, and the site number was painted on the roof so pilots and ground crews could confirm their position at a glance. This rigorous physical network layer highlights how the engineering of trust through measurement systems underpins all institutional mechanics before automated communication lines exist. Some of these outbuildings doubled as weather-observation stations, folding a second layer of aviation infrastructure into what looked, from a distance, like a single lonely shack in the desert.

By 1933, the system’s high-water mark, the Transcontinental Airway System comprised approximately 1,500 beacons across roughly 18,000 miles of lighted airway, alongside 236 lighted emergency landing fields spaced every 15 to 20 miles for pilots who needed to put a plane down in a hurry. That is a staggering amount of federal infrastructure to build in under a decade, using construction crews, generator technicians, and maintenance staff scattered across some of the most remote terrain in the country, all to solve a problem that would be rendered largely obsolete within another ten years.

None of this reads the same way once you imagine actually sitting in the cockpit. Airmail pilots in the 1920s flew open-cockpit biplanes like the de Havilland DH-4, a design so prone to catching fire in a crash that pilots grimly called it the “Flaming Coffin.” There was no heater, no pressurization, and in the earliest years, no radio worth trusting. A pilot dressed for the flight the way a person might dress for standing outside in a blizzard for six hours, because that’s essentially what the job required.

Following the arrows was, in principle, close to mechanical: keep the current arrow lined up ahead of you, and when you passed it, look for the next one along the indicated heading. In practice, it demanded constant vigilance. Arrows were easy to lose track of in haze, in the long shadows of early morning or late afternoon, or simply because a pilot flying at a sensible cruising altitude was sometimes too high to make out a 70-foot painted shape against miles of open desert. At night, obviously, the arrows themselves were invisible, and pilots depended entirely on the rotating beacons and their course lights to confirm they hadn’t drifted off route.

The weather made all of this worse in ways that are easy to understate from the safety of a desk chair. Fog could erase a beacon’s light entirely. Mountain passes in Wyoming, Utah, and Nevada created turbulence and downdrafts that could push an aircraft into terrain before a pilot had time to react. Winter storms across the Great Plains, the same storms Jack Knight flew through in 1921, remained a constant hazard even after the beacon system matured, because a rotating light is only useful if you can actually see it through the snow. Near-disasters were routine enough that they rarely made the newspapers unless a pilot died; surviving one simply meant you flew the next scheduled leg.

It’s worth being precise here about who did and did not fly this particular system. The 1910s produced a generation of extraordinary American aviators who normalized the idea that flying was worth risking your life for, pilots like Ruth Law, who set cross-country distance records, delivered the first official U.S. airmail to the Philippines in 1919, and spent the era proving, loudly and publicly, that women belonged in the cockpit. Law retired from flying in 1922, a year before Congress funded the first segment of the lighted airway, so she was never part of the beacon and arrow network described in this article. But the culture of daring, public, high-stakes flying that pilots like her built in the preceding decade is part of why the country was willing to trust airmail pilots with a multi-thousand-mile chain of navigational infrastructure in the first place.

Night navigation deserves its own accounting, because it was the entire reason the system existed. A concrete arrow is useless after sunset. Everything the Department of Commerce built for nighttime flying depended on light: the rotating beacons on their towers, the fixed course lights aimed at neighboring stations, and, at the emergency landing fields spaced every 15 to 20 miles along the route, floodlights powerful enough to illuminate a usable landing strip in open country with no other lighting for miles.

This is where the system’s ambition becomes clearest. Building a chain of towers is one thing. Building a chain of towers, backup generators, weather shacks, and emergency airfields spaced closely enough that a pilot flying at roughly 90 miles per hour would encounter some kind of navigational or safety infrastructure every six to twelve minutes, across nearly the entire continental United States, is closer to building a second, invisible interstate highway system, a full generation before the interstate highway system itself existed. It required the same immense logistical foresight as Roman harbor engineering, adapting massive, hostile environments for safe transit operations.

The Transcontinental Airway System’s obsolescence was baked into its own success. Radio navigation technology, developing in parallel throughout the 1920s, matured rapidly in the following decade. Low-frequency radio range systems, introduced starting around 1929, gave pilots a way to navigate using radio signals rather than a physical chain of landmarks, a solution that worked in fog, worked at night without a visible light source, and didn’t require a technician to keep a generator running in a shed in the middle of Nevada.

The Great Depression accelerated the transition by making the visual system’s operating costs, generator fuel, maintenance crews, replacement equipment scattered across thousands of miles of remote terrain, increasingly hard to justify next to a cheaper radio-based alternative. The last new concrete arrow was reportedly built in 1931. Through the 1930s, the Department of Commerce steadily decommissioned towers as radio range stations came online to replace them.

World War II delivered the final blow twice over. Steel was desperately needed for the war effort, so most of the remaining beacon towers were dismantled and scrapped. And in a strange, almost paranoid postscript, officials in some areas deliberately destroyed or obscured surviving arrow markings out of concern that enemy aircraft might use the same navigational aids American pilots had relied on for two decades. This logic of deliberate legibility and illegibility echoes why addresses have numbers, shifting based on security and institutional control. The system built to make American skies legible was, for a brief period, actively made illegible again, on purpose, as a wartime precaution.

A handful of beacons soldiered on afterward. Montana kept 19 of the original towers operational, maintained by the state’s Department of Transportation Aviation Division, into the 21st century, a remarkable stretch of continuous service for 1920s infrastructure. The Federal Aviation Administration decommissioned its last operating beacon in the early 1970s. But by the early 1940s, for most of the country, the age of navigating by concrete arrow was already over.

Here is the detail that turns this from a history lesson into something you can actually go and look at: the towers were valuable enough to strip for scrap, but the arrows were poured concrete, embedded in the ground, and worth almost nothing to salvage. Nobody bothered removing most of them. They’re still there.

Hundreds likely survive, although no official nationwide census exists. Dedicated enthusiasts, genuine amateur historians running websites like Arrows Across America, off-roading forums, and local historical societies, have documented dozens of intact arrows across more than a dozen states, concentrated heavily in the remote deserts of the Southwest and Mountain West where there was never enough development pressure to bother clearing them. This forgotten network remains one of the most fascinating physical artifacts in aviation navigation history.

Google Earth and satellite imagery have quietly turned amateur arrow-hunting into a small but genuine subculture. Spotting these concrete arrows via Google Earth coordinates has become a way to digitally explore 1920s federal infrastructure. Hikers and off-roaders occasionally photograph sites nobody has formally documented before, and forum threads devoted to tracking down coordinates read like a slow-motion, crowdsourced archaeological survey of early federal infrastructure, conducted by hobbyists rather than historians, largely because no government agency ever felt obligated to preserve what it built. If you are looking for a giant concrete arrows map, none officially exists, but amateur efforts continue to uncover new coordinates every year.

Why This Still Matters

Step back from the arrows themselves and a larger pattern comes into focus. America has a habit of solving enormous logistical problems by rewriting the landscape itself, then quietly forgetting it did so. Long before this, surveyors had already imposed mathematical order on the continent via the Jeffersonian Grid, but the airway took this into the sky. Railroads eventually forced the entire country onto standardized time just to keep their schedules coherent, a primary reason why time zones were created. And for about two decades, before radio made it unnecessary, the federal government turned the ground itself into a readable instrument, painting directional information directly onto the earth so a person a thousand feet in the air could look down and know, unambiguously, where to go next.

That’s a genuinely clever solution to a genuinely hard problem, and it sits in an interesting lineage. Roman roads used milestones and fixed markers to organize movement across an empire. Lighthouses used light and position to warn ships away from danger and toward safe harbor. GPS satellites now do essentially the same job the arrows did, telling you where you are and where to go, using signals instead of concrete. The arrows are a strange middle chapter in that same long story: navigation solved not with electronics, but with geometry, paint, forgotten surveying tools, and a very large amount of poured concrete, built by a government that needed pilots to survive long enough for something better to come along.

Forensic Archive: Primary Sources

The institutional records, technical bulletins, and pilot accounts used to verify this story:

• U.S. Bureau of Air Commerce, Airway Bulletin No. 1 (September 1, 1932): Original technical diagrams and specifications for beacon and arrow construction.
• National Trust for Historic Preservation: “The True Story Behind Those Giant Concrete Arrows,” including interviews with National Postal Museum historians.
• Library of Congress: Historic newspaper archives verifying Jack Knight’s contemporary coverage and flight timelines.
• National Postal Museum, Smithsonian Institution: Pilot biographies and primary documents for Jack Knight and the 1921 relay flight.
• National Air and Space Museum, Smithsonian Institution: Ruth Law biographical archive.
• Montana Department of Transportation, Aviation Division: Records on the state’s continued maintenance of original beacon towers.
• Federal Aviation Administration historical archives: Decommissioning records for the last operating beacons.

Back Over Wyoming

Our pilot from the opening scene reaches the next beacon exactly where the chart said it would be, banks slightly, and keeps flying toward San Francisco. He has no idea that in fifteen years his entire system will be scrapped for its steel, or that the one thing nobody bothers to remove, a slab of yellow-painted concrete, will still be sitting in that same patch of desert nearly a century later, waiting for a hiker to find it and wonder what on earth it’s doing there. Today, satellite imagery has made these forgotten arrows easier to find than at any point since they guided pilots nearly a century ago. He just adjusts his heading a few degrees, the way he has a hundred times before, and keeps going.