The Mw 4.6 Odaesan, Korea, earthquake of 20 January 2007 (11 hr 56 min) had nine locatable foreshocks in the previous 67 hours and nine aftershocks during the 11 hours after the main event. We obtained accurate locations of this sequence by employing the double-difference earthquake relocation method with differential travel-time measurements using waveform cross-correlation. We found that 19 accurately relocated foreshocks, mainshock, and aftershocks aligned along a 1.2 km long lineation with an azimuth of west-northwest–east-southeast (115°), which coincides with a nodal plane of the mainshock focal mechanism that strikes 114°. Regional waveform modeling of the mainshock constrained a focal depth of 11±1 km and the focal mechanism is a pure strike-slip faulting along the vertical nodal planes with a horizontal P axis (plunge=1°) trending east-northeast (69°). The pulse rise time of the mainshock source time function estimated from an empirical Green’s function analysis ranges from 0.1 to 0.16 sec, which yields a source radius of ∼0.6 km and is consistent with the spatial distribution of the foreshock and aftershock sequences. The stress drop of is obtained for the mainshock using a radius of 0.6 km for a circular rupture, and seismic moment . The 20 January 2007, Odaesan earthquake may be the first significant earthquake in southern Korea to date where the fault plane is constrained by accurate locations of the foreshock and aftershock sequences. Its source mechanism, vertical strike-slip fault at depth, suggests that the event represents reactivation of a buried high-angle fault in the Precambrian basement by the contemporary east–east-northeast trending regional horizontal compressive stress.