A joint hypocenter determination routine has been developed to accurately determine the locations of earthquakes occurring under Mount St. Helens. The method has a special application to Mount St. Helens, where most of the earthquakes are suspected to occur at shallow depth in a rather limited volume under the volcano. Differences in elevation among stations are allowed for in this routine. The location procedure determines a best half-space velocity and one station correction for each station simultaneously with earthquake hypocenters. Arrival time data from an array of portable digital seismic event recorders are supplemented with data from stations of the permanent University of Washington seismic network operating at the volcano. Earthquakes are found to occur along a line trending southeast of the dome inside the crater. The shallowest events occur near the dome with hypocentral depth increasing with increasing distance from the dome. Station corrections can be interpreted to imply the existence of a high-velocity body beneath the crater which extends at least between depends of .5 and 2.1 km beneath the crater floor. Spectra are calculated from earthquake waveforms recorded at three stations. The spectra show very little evidence for local site or path effects on frequency content of waveforms. Stress drops of earthquake are calculated from measurements of corner frequency and low-frequency spectral amplitude. Stress drops of earthquakes vary between .5 and 5 bars. There is no evidence for a limiting corner frequency for small earthquakes as has been found for earthquakes occurring in Mammoth Lakes, California. Data fit well a liner trend between log seismic moment and log source radius fit to data for earthquakes ranging in size from 1014 to 1027 dyn cm, suggesting that there is no limiting source radius observed for earthquakes in this range of size.