Weather station barograph records as well as infrasonic recordings of the pressure wave from the Mount St. Helens eruption of May 18, 1980, have been used to estimate an equivalent explosion airblast yield for this event. Pressure amplitude versus distance patterns in various directions compared with patterns from other large explosions, such as atmospheric nuclear tests, the Krakatoa eruption, and the Tunguska comet impact, indicate that the wave came from an explosion equivalent of a few megatons of TNT. The extent of tree blowdown is considerably greater than could be expected from such an explosion, and the observed forest damage is attributed to outflow of volcanic material. The pressure-time signature obtained at Toledo, Washington, showed a long, 13-min duration negative phase as well as a second, hour-long compression phase, both probably caused by ejacta dynamics rather than standard explosion wave phenomenology. The peculiar audibility pattern, with the blast being heard only at ranges beyond about 100 km, is explicable by finite amplitude propagation effects. Near the source, compression was slow, taking more than a second but probably less than 5 s, so that it went unnoticed by human ears and susceptible buildings were not damaged. There was no damage as Toledo (54 km), where the recorded amplitude would have broken windows with a fast compression. An explanation is that wave emissions at high elevation angles traveled to the upper stratosphere, where low ambient air pressures caused this energetic pressure oscillation to form a shock wave with rapid, nearly instantaneous compression. Atmospheric refraction then returned part of this wave to ground level at long ranges, where the fast compressions were clearly audible. ¿ American Geophysical Union 1987