Short answer
The U.S. Geological Survey states directly that Yellowstone is not overdue. Its major caldera-forming eruptions occurred roughly 2.08 million, 1.3 million, and 631,000 years ago. Those dates provide only two intervals, and the intervals are not equal. Averaging them does not turn the volcanic system into a clock. [1]
Where the countdown comes from
A common version of the claim says Yellowstone produces a very large eruption about every 600,000 years and that the last one occurred more than 600,000 years ago. The arithmetic sounds intuitive because it resembles a maintenance interval. But a volcano is not a machine with a scheduled service date.
Even the arithmetic is misrepresented. The two observed gaps are of different lengths, and their average is longer than 600,000 years. More importantly, two intervals are far too little evidence from which to infer a periodic process. The average describes those two gaps; it does not establish the next one. [2]
Not overdue does not mean inactive
Rejecting the countdown is not the same as declaring Yellowstone harmless or extinct. It remains an active volcanic and hydrothermal system, and another eruption is possible. The point is narrower: past caldera-forming eruptions do not provide a reliable timetable for the next event.
Risk communication fails when it forces a choice between “a supereruption is imminent” and “nothing can happen.” Scientists can monitor earthquakes, ground deformation, thermal features, and gas without claiming a date the evidence cannot support.
A future eruption need not be a supereruption
The public image of Yellowstone often jumps directly from unrest to a continent-scale caldera event. The geological record is broader. USGS notes that most Yellowstone eruptions were not highly explosive caldera-forming events; lava flows are more common in the record than supereruptions. [3]
That does not make smaller volcanic or hydrothermal events trivial. It means the hazard should be described as a distribution of possible events rather than a single cinematic scenario.
The statistical mistake
An average recurrence interval is useful only when the process and the evidence justify treating events as repeated observations from a stable pattern. Here, the sample is two intervals from a changing geological system. The value has no demonstrated power to forecast the next caldera-forming eruption.
Calling the system overdue adds a second unsupported assumption: that the probability rises sharply after the average interval passes. The observed dates do not establish such a threshold. A tail risk can be real without being periodic.
The decision boundary
Supported
Yellowstone has produced major eruptions, remains an active monitored system, and could erupt again.
Not supported
The three major eruption dates do not establish a deadline, a periodic schedule, or an imminent supereruption.
What to use instead of a countdown
Use the current assessments of the Yellowstone Volcano Observatory and local emergency authorities, not a recycled average. Monitoring evaluates the system that exists now. A recurrence myth substitutes an appealing story for that live evidence.
Chapter 10 of The Volcanic Engine uses this case to make a broader argument: uncertainty becomes more useful when it is stated as uncertainty. A sober hazard is not made safer by exaggeration, and a correction is not a promise that nothing will happen.