During May and early June 1979, five three-component digital seismographs were deployed near the Monticello Reservoir, South Carolina, to determine the stress drops of reservoir-induced earthquakes and compare them with earlier hydrofracture stress measurements of Zoback and Hickman (this issue). Records for 327 events were obtained; 10 of these events, each recorded on four or more stations, were analyzed to determine stress drops. Depths for the 10 events ranged from 0.5 to 1.4 km. Fault plane orientations were calculated from an inversion of the body wave displacements; individual radiation pattern corrections were applied in calculating the moments at each station. Final estimates of seismic moment were obtained from the individual estimates of the source radius, and stress drop was then calculated once from the final estimates of seismic moment and source radius for each event. Moments of the 10 digitally recorded earthquakes ranged from 2.5¿1017 to 2.7¿1018 dyne cm for the P wave data and from 4.6¿1017 to 4.6¿1018 dyne cm for the S wave data. The stress drops (S wave data only) generally ranged from about 0.2 to 4 bars. Four earthquakes were recorded by a strong motion instrument at the dam site. Magnitudes (M) for these events ranged from 2.8 to 3.0. Moments taken as averages of the values from each component ranged from 1.7 to 3.4¿1020 dyne cm and the stress drops from 13 to 92 bars. Large differences between the corner frequencies of the two components (factors of 2.5 and 4 for two of the events) correspond to a large relative uncertainty for these stress drop values. Nevertheless, tens of bars of stress are apparently released for the largest events in the Monticello sequence but only a few bars are released for the smaller magnitude events. In comparison the in situ stress data measured in two holes give an estimate of shear stress of about 45 bars at 0.2-km depth and 30 to 70 bars at 1-km depth in the northern hole and from 15--20 bars near the surface to almost 70 bars at depth in the sourthern hole. The variation in shear stress inferred from in situ measurements at the deeper levels indicates an inconsistency between the thrust-faulting mechanisms of the earthquakes and the state of stress which is more consistent with thrust faulting at 0.2-km depth and strike slip or normal faulting below. Thus the estimates of shear stress from in situ measurements of the same order as the stress drops of the M≂2.8-3.0 events but about 1 order of magnitude larger than the stress drops obtained for the smaller events.