Short-period teleseismic and near-regional long-period waveforms from a large aftershock (mb = 5.3, origin time 06:18:36.8 UTC) of the 10 December 1967 Koyna main event were modeled to determine source depth and focal mechanism. Identification of pP and sP on the short-period waveforms yielded a well-constrained source depth of 3.5 to 4.0 km. The focal mechanism was determined using a systematic trial and error (grid-testing) technique in which seven P first motions, two SH first motions, two teleseismic pP/P ratios, one teleseismic sP/P ratio, and one regional SH/P ratio were included. This heterogeneous, but sparse, data set yielded a constrained normal fault mechanism at 100° ± 20° strike, 40° ± 10° dip, and 240° ± 20° rake. Rupture directivity is inferred from consistently high surface reflection/direct wave amplitude ratios. A circular fault of 1 km radius with a rupture initiation point at the lower edge and an assumed rupture velocity of 0.9 Vs explains the anomalously high amplitude ratios and yields a seismic moment of 3.8 × 1023 dyne-cm. The shallow source depth is similar to that found for the main event (4.5 km) and the 13 September foreshocks (5 km) in previous modeling studies and is consistent with the premise that the Koyna earthquakes were triggered by the impoundment of the Koyna reservoir. However, the normal fault mechanism differs substantially from the previous events which were left-lateral strike-slip faults trending NNE. The differing aftershock mechanism is probably due to the reactivation of old faults in the Koyna area caused by stress readjustment after the Koyna main event.