In the winter of 1811 and 1812, a sequence of earthquakes beneath the Mississippi River valley made the river appear to run backward near New Madrid, Missouri, toppled chimneys in St. Louis and rang church bells more than 700 miles away in Richmond, Virginia. The U.S. Geological Survey now puts the odds of a comparably sized earthquake striking the same ground at 7 to 10 percent within the next 50 years, a hazard that spans seven states: Illinois, Missouri, Arkansas, Kentucky, Tennessee, Mississippi and Indiana.
A fault system with no plate boundary underneath it
Unlike the San Andreas Fault in California, the New Madrid Seismic Zone does not sit along the edge of two tectonic plates grinding past each other. It runs through the interior of the North American plate, tracing an ancient rift more than 500 million years old that never fully healed after the continent tried, and failed, to split apart in that spot. Regional stress reactivates that old weakness today, which is a large part of why researchers still cannot fully explain what triggers individual New Madrid earthquakes on any predictable schedule.
Seismometers have recorded the zone’s activity since 1974, logging thousands of small to moderate earthquakes each year along faults buried thousands of feet beneath farmland in the Mississippi embayment. The thick, water-saturated sediment there carries shaking far beyond where a similarly sized earthquake would travel through the harder bedrock of the West Coast, which helps explain why the 1811-1812 sequence swayed furniture and sent residents fleeing their homes as far away as Ohio and Indiana, hundreds of miles from where the ground actually broke.
Seven states inside the shaking radius
Geologic mapping of where the 1811-1812 quakes caused the heaviest damage puts Illinois, Missouri, Arkansas, Kentucky, Tennessee, Mississippi and Indiana inside the seismic zone’s reach. Memphis, Tennessee, sits closest to the zone’s southern end and would likely see the strongest shaking of any major city in a repeat event; St. Louis, Missouri, sits near its northern edge and had chimneys knocked down in 1812 despite being roughly 150 miles from the epicenter. Smaller towns in northeastern Arkansas and southwestern Kentucky occupy the zone’s core, where shaking-driven liquefaction destroyed the original settlement of Little Prairie outright.
The earthquakes themselves arrived as a rapid cluster rather than a single event: a magnitude-7.5 shock before dawn on December 16, 1811, a second major jolt hours later that same day, a magnitude-7.3 rupture on January 23, 1812, and a final magnitude-7.5 quake on February 7, 1812, that leveled the town of New Madrid and dropped a stretch of ground in Tennessee low enough for the Mississippi River to pour in and form Reelfoot Lake.
The number the U.S. Geological Survey keeps repeating: 7 to 10 percent
Geologist David Applegate of the U.S. Geological Survey told a 2007 congressional hearing that agency scientists calculate “a 7 to 10 percent chance of a New Madrid earthquake the size of those in 1811-12 occurring in the next 50 years,” alongside a considerably higher 25 to 40 percent chance of a smaller, magnitude-6 earthquake in the same window. A separate USGS fact sheet arrives at the same 7-to-10-percent figure independently, using paleoseismic evidence dug out of sand blows and liquefaction scars that older, prehistoric earthquakes left buried across the region.
That estimate has held roughly steady across two decades of follow-up research, even as scientists have argued over whether the zone is winding down or recharging. Modern GPS measurements show comparatively little strain building along the mapped faults today, a puzzle that has fueled a running debate over whether 1811-1812 was a rare, one-off release of ancient stress or the latest turn in a hazard that keeps reloading on a centuries-long cycle.
Buildings raised for wind, not for shaking ground
Applegate’s testimony also flagged what he called a preparedness gap between the two U.S. coasts most associated with earthquakes and the interior states that actually hosted the country’s largest historical sequence. “The impacted communities are less prepared than their west coast counterparts,” he said. “Most of the buildings and infrastructure were built without consideration of seismic shaking.” Unreinforced brick storefronts, older school buildings and highway bridges across the seven states were largely engineered against tornadoes, river flooding and heavy snow loads, not the kind of ground motion New Madrid produced twice within six weeks two centuries ago.
The Central United States Earthquake Consortium, a coalition formed by the states inside the seismic zone, coordinates the retrofit programs and emergency drills meant to close that gap. Its coverage map tracks building-code updates county by county rather than state by state, since the hazard crosses jurisdictions that otherwise rarely plan for a shared disaster together.
The U.S. Geological Survey’s own program page on the zone is candid about the limits of that planning. The agency states plainly that “no one knows what causes New Madrid earthquakes,” even as the instruments installed since 1974 keep recording the small tremors that make the interior fault system one of the most closely watched in the country east of the Rocky Mountains.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
