Trench excavation is a less effective paleoseismic investigation for intramontane active faults because there are fewer good trench sites along them. We propose a new method called “pit excavation on a mountain slope” to determine the timing of the latest events of such faults. This method is carried out across a small uphill-facing fault scarp preserved on a mountain slope. Sediment trapped by the scarp should provide good information concerning the timing of the latest event.<BR>We applied the method to the northern segment of the Daguchi fault, one of the short rightlateral active faults in the Nosaka Mountains, north of Lake Biwa, southwest Japan. The rupture history of the segment has remained unknown because no good trench site exists along the fault trace. However, our precise geomorphological mapping revealed the presence of a 1 m-high uphill-facing fault scarp preserved on a mountain slope, and therefore we carried out a pit excavation across the scarp using man power. A very clear fault and 75 cm-thick trapped sediments were observed on the pit wall. Since these sediments showed neither displacement nor deformation associated with surface faulting, we interpret that they were all deposited after the latest event. For more accurate age control, we identified species and portion for wood fragments from the sediments. Unfortunately, all samples except those that could not be identified were parts of Cryptomeria japonica, Chamaecyparis obtusa Endl. or Cupressaceae, which can live for hundreds or even thousands of years : this means a possible large offset between deposition and radiocarbon age for these samples, leading to less accurate age control. However, some of them were from the outermost trunk, for which we can regard radiocarbon age as the deposition age in consideration of the very small catchment area of the pit site (several hundreds of m²). Based on the radiocarbon ages of the outermost trunk samples, the initiation of sediment trapping or equivalently the timing of the latest event is estimated to be about 1500 years ago (5th to 7th century AD).<BR>Our investigation on the Daguchi fault successfully revealed the accurate timing of the latest event, supporting the effectiveness of our new method. However, some problems remain unsolved, including application of the method to other faults, especially those with long dormancy, and necessity of a crosschecking excavation on a historically ruptured fault.