In general, no single event can be used to predict a volcanic eruption, and thus many events are usually monitored so that taken in total, an eruption can often be predicted. Still, each volcano behaves somewhat differently, and until patterns are recognized for an individual volcano, predictions vary in their reliability. Furthermore, sometimes a volcano can erupt with no precursor events at all.

After the catastrophic eruption of Mount St. Helens on May 18, 1980, a volcanic dome began to grow in the crater. Growth of this dome occurred sporadically, and sometimes small eruptions occurred from the dome. After several years of dedicated monitoring, scientists are now able to predict with increasing accuracy eruptions from this dome. An example is shown in the graphs to the right. In the weeks prior to an eruption on March 19, 1982, the amount of seismic energy released increased, the amount of dome expansion increased, tilt increased, and SO2 emissions increased prior to the event.

Note that eruption predictions such as in this example are only possible if constant monitoring of a volcano takes place. Monitoring is an expensive endeavor, and not all active or potentially active volcanoes are monitored. Still, if people living around volcanoes are aware of some of the precursor phenomena that occur, they may be able to communicate their findings of anomalous events to scientists who can begin monitoring on a regular basis and help prevent a pending disaster.