On the morning of June 5, 1976, a bulldozer working beside the growing leak on Teton Dam’s downstream face began sinking into the embankment itself.

On the morning of June 5, 1976, workers at Teton Dam in southeastern Idaho were trying to stop a hole that had appeared on the downstream face of a nearly completed federal dam. The hole was carrying muddy water, which was already a worse sign than clear seepage because it meant the water was taking pieces of the embankment with it.

They brought in heavy equipment and began pushing rock toward the opening.

One operator later described the machine settling as the ground beneath it disappeared. Reclamation’s present-day reconstruction says two bulldozers were lost to the enlarging hole around 11:20 that morning. Both operators escaped. One machine was never recovered.

At 11:57, the dam broke.

The bulldozers had gone in first.

The dam was almost new

Teton Dam had been authorized by Congress in 1964 and construction began in February 1972. It was an earth-and-rock embankment built across the Teton River northeast of Rexburg, intended to provide irrigation storage, hydroelectric power, flood control and related water benefits to the upper Snake River region. The structure rose about 305 feet and stretched roughly 3,100 feet across the canyon.

By November 1975, the dam itself was essentially complete.

The reservoir had already begun filling.

That distinction matters because Teton failed during its first filling, before the project had settled into ordinary operation. The two planned generating units were installed but had not yet entered normal service. Some outlet and pumping facilities were still unfinished.

The site had always presented an unpleasant geological problem.

The canyon walls were volcanic rock, extensively fractured and locally porous. Engineers knew water could move through those fractures, so construction crews drilled and injected grout to create a barrier beneath and beside the dam. The work consumed far more grout than initially expected because the rock contained fissures that were wider and more numerous than anticipated.

The unusually high grout consumption showed how fractured the foundation rock was. On June 5, reservoir water exploited remaining pathways near the right abutment.

The easy explanation does not quite work

The reservoir was filling quickly during the spring of 1976.

A normal filling guideline of about one foot per day had been relaxed in March. Heavy snowmelt then pushed rates higher, averaging around three feet per day at times and briefly reaching more than four. That has made “they filled the dam too fast” one of the enduring shorthand explanations for Teton.

The official investigations did not find that satisfactory.

Idaho’s Office of Emergency Management, summarizing the post-failure reviews, notes that investigators examined the accelerated filling closely and did not identify it as the cause of the collapse.

The harder explanation involved the structure itself.

Investigators focused on the interaction between the dam’s fine-grained core, the highly fractured volcanic foundation and abutment rock, and the pathways by which reservoir water could reach places where it began carrying soil away. The precise initiating mechanism remained debated among the various investigations, but internal erosion around the right abutment was central to every serious reconstruction.

The independent review went farther than saying something unforeseeable had happened.

Its conclusion, as summarized by Idaho emergency-management history, was that a safe dam could have been built at the site using design concepts already known at the time.

That sentence is less comforting than blaming unusual geology.

June 3

The first visible warnings were not dramatic.

On June 3, inspectors found clear water emerging downstream of the dam in two locations. Another seep appeared on June 4.

Clear seepage through or around an embankment dam is not automatically evidence of imminent failure. The observations were recorded, but none was considered a serious hazard at the time.

Early on June 5, that changed.

Between roughly seven and eight in the morning, muddy water was discovered near the downstream toe. More seepage appeared.

Mud means transport.

Water was no longer merely finding a route.

It was bringing the dam with it.

Around ten o’clock, a much larger leak developed near the right abutment.

By 10:30, witnesses described a roar like a waterfall.

Jerry Dursteller arrived around that time with a camera.

He kept taking pictures.

That turned out to be unusually important.

A sequence no model had to recreate

The National Archives preserves Dursteller’s photographs among Bureau of Reclamation records.

They are not merely before-and-after photographs.

They are a sequence.

One frame shows a muddy stream cutting into the downstream face.

The next shows the cavity getting larger.

Another shows the hole chewing upward through the embankment.

His final photograph, made shortly before he ran out of film around 11:50, shows the cavity reaching the crest.

The remarkable thing is how little interpretation the sequence requires.

The dam is visibly disappearing.

Meanwhile, a different sign appeared on the reservoir side.

A whirlpool.

Witnesses reported it beginning only a few feet across and rapidly growing to perhaps twenty feet in diameter. Water was being drawn into whatever passage had opened through the right side of the dam.

The engineers now had both ends of the problem.

A whirlpool where reservoir water was going in.

A muddy torrent where pieces of the dam were coming out.

Bring the bulldozers

Workers tried the most immediate thing available.

Push material into the hole.

Heavy equipment began moving rock and earth toward the eroding section. The attempt was crude only in hindsight. Internal erosion was carrying material out faster than the operators could replace it.

The ground around the machines began to fail.

Reclamation’s current timeline places the loss of two bulldozers at about 11:20. The operators escaped as the cavity enlarged. One machine was never recovered.

The attempt was meant to slow the erosion by pushing rock into the growing cavity and buy time for another response. The ground failed faster than the material could be placed.

If the flow could be slowed, perhaps the erosion could be arrested long enough for another response.

The dam did not give them that interval.

10:43

While people at the site were trying to save the structure, another operation had already begun.

At 10:43 a.m., project officials warned local sheriffs that the dam might fail and downstream communities might need evacuation. The warning was soon upgraded.

Get people out.

Madison and Fremont County officers began calling homes they knew lay in the flood path and driving through threatened communities telling residents to leave. Local radio joined the warning effort. Idaho’s emergency-management history credits that improvised combination of telephone calls, officers, local officials and word of mouth with evacuating threatened towns quickly enough to prevent a far larger death toll.

The National Archives preserves the transcript of a live KRXK broadcast from the dam.

Reporter Don Ellis was watching the structure come apart while telling people downstream, in increasingly plain language, to get out.

The chronology shows the contrast plainly: while the federal dam was failing, local sheriffs, radio, telephone calls and word of mouth were already moving people out of the flood path.

11:57

At 11:55, the crest began to sag.

Two minutes later, the right portion of Teton Dam gave way.

More than 250,000 acre-feet of reservoir water—over eighty billion gallons—began rushing into the canyon below. Reclamation places peak discharge around one million cubic feet per second. The reservoir, about 270 feet deep at the dam before failure, drained in less than six hours.

The first communities had very little time.

Wilford was struck first.

Newdale, Teton and Sugar City followed. Around one in the afternoon, Sugar City encountered floodwater reported at roughly fifteen feet high. Rexburg, the largest community directly in the path, was next.

The flood picked up lumber, equipment, vehicles, agricultural debris and pieces of buildings.

Anything that floated became part of the water.

Anything that did not float had to resist it.

Much of it did not.

Eleven

The Teton flood eventually affected roughly 300 square miles and up to 100,000 acres of agricultural land. Hundreds of homes were destroyed, thousands more were damaged, businesses were wrecked, livestock losses ran into the tens of thousands, and transportation and utility systems were torn apart.

Eleven people died.

The number is startlingly low compared with the scale of the flood.

That does not make eleven small.

It means the warnings mattered.

The flood was not survivable everywhere it went, but most people reached higher ground before it arrived. Ricks College in Rexburg sat above much of the inundated area and opened dormitories and other facilities to displaced residents. Local churches, volunteers and government agencies built a relief operation around the hill while muddy water still occupied large portions of town.

One of the more revealing photographs from the aftermath looks north over flooded Rexburg.

The college sits above it.

Dry.

A shelter because somebody had built it in the right place long before anybody knew it would need to become one.

The investigators dug into what remained

A catastrophe this visible encourages a simple question.

What broke?

Teton produced a less simple answer.

Multiple investigative groups examined the dam, foundation, grout curtain, fill material, construction records and eyewitness evidence. The National Archives now holds reports, depositions, photographs and hearing material from years of that work.

The broad picture is clear.

Reservoir water found a route into the right abutment area.

The highly fractured volcanic rock and the way the dam’s impervious core interacted with that foundation allowed internal erosion to begin.

Once material started moving, the process accelerated until the embankment could no longer support itself.

Exactly how the initial hydraulic fracture or erosion path formed remained less certain.

That distinction matters.

“Teton failed because engineers ignored cracked rock” is too neat.

They knew the rock was fractured.

They grouted it.

The failure came from the fact that the design and treatment did not adequately control what reservoir water could do once it found the remaining weaknesses.

The site had been studied.

The wrong thing still happened inside it.

The dam had a predecessor

There was an especially uncomfortable precedent.

In 1965, Reclamation’s Fontenelle Dam in Wyoming developed serious seepage and a major internal cavity while its reservoir was being filled. Crews lowered the reservoir, worked continuously on the damaged area, poured enormous quantities of grout into the foundation and prevented a breach.

Teton was built afterward.

The two dams were not identical, and it would be sloppy to say Reclamation simply repeated the same design and received the same result.

But Reclamation itself now discusses the Fontenelle near-failure and Teton together when explaining the development of its modern dam-safety approach.

At Fontenelle, the outlet works allowed the reservoir to be drawn down fast enough to buy time.

At Teton, the main outlet works were not yet operational when the failure developed.

The difference between a frightening repair story and a catastrophe can be surprisingly mechanical.

How quickly can you get the water out?

After Teton

Reclamation did not rebuild the dam.

Congress instead passed the Reclamation Safety of Dams Act in 1978, and Teton became the catalyst for the agency’s modern Dam Safety Program. Existing dams were subjected to more systematic review, with greater emphasis on deficiencies that might not have been understood when the structures were originally built.

That program still exists.

Reclamation describes its modern process in terms of repeated inspection, updated engineering analysis, risk assessment and corrective action when a structure presents unacceptable risk to people, property or the environment.

A useful comparison is Inside the Joint, the Silver Bridge investigation. Silver Bridge failed around a crack hidden inside a critical joint; Teton failed through internal erosion that remained largely invisible until muddy water appeared outside the dam. Both disasters changed what inspection had to accomplish: seeing a structure was no longer enough to understand what was happening inside it.

The machine

The remains of Teton Dam are still in the canyon.

The river passes through what is left.

In June 2026, fifty years after the failure, the Bureau of Reclamation opened a new overlook at the site with interpretive panels describing the construction, geology, collapse, downstream disaster and dam-safety program that followed.

One panel is a timeline.

June 5.

Seven to eight in the morning: muddy seepage.

Ten: a large leak.

10:43: warn the sheriffs.

11:20: two bulldozers lost.

11:30: another hole.

11:57: breach.

That is the federal agency’s own version now.

A dam whose purpose included flood control created the flood.

The operators lived.

One bulldozer never came back.

Sources & Further Reading

• Bureau of Reclamation, Teton Dam History. Reclamation’s current institutional history, including the failure timeline, construction geology, disaster effects and dam-safety legacy.
• Independent Panel to Review Cause of Teton Dam Failure, 1976 report, and Department of the Interior Teton Dam Failure Review Group final report. The two principal federal investigations into how the structure failed.
• National Archives at Denver, Record Group 115, Bureau of Reclamation Teton Dam failure records. Includes Jerry Dursteller’s near-continuous photographic sequence, reservoir records, investigation documents and post-failure evidence.
• Idaho Office of Emergency Management, “1976 Teton Dam Collapse.” Particularly useful for reservoir-filling history, geology, warning and evacuation, downstream losses and later interpretation.
• U.S. Geological Survey, Teton Dam Flood Hydrologic Atlas series. Field-mapped inundation and high-water evidence extending roughly 100 miles downstream.
• Bureau of Reclamation Dam Safety Office. Current institutional account of Teton’s role in the Reclamation Safety of Dams Act and modern risk-based safety program.