Terrestrial paleoclimatic and paleoenvironmental records on the Indian monsoon are extremely limited, though many investigations have been done on the deep-sea sediments in the Indian Ocean during the last few decades. In order to clarify the terrestrial monsoonal climatic records during the Quaternary, we undertook core-drilling of the basin-fill sediments of the Kathmandu Valley, which is located under Indian monsoon zone. In this paper, we attempt to reconstruct the environmental changes of the Pakeo-Kathmandu Lake, on the basis of studies on the fossil-diatom collected from a 218-m-long core drilled at Rabibhawan, western central part of Kathmandu. Except the uppermost thin cover of fluvial sediments (Patan Formation), the core is continuous and mainly composed of muddy lacustrine sediments (Kalimati Formation) containing abundant fossil-diatoms, which document ecological responses to climatic and environmental changes (Sakai 2001, Hayashi et al. 2002). Mud samples were collected at 50 cm interval from 7 to 45.5 m in depth and at 2 m interval from 47.5 to 218 m in depth. We identified each species and counted the number of valves by means of a scanning electron microscope (JSM-5600) at x5,000 magnification, because dominant species in the RBcore are very small, ranging from 5 to 30 μm. For each sample, at least 300 diatom valves were counted. On the basis of relative abundance and the number of principal diatoms, eight fossil zones and three subzones in zone 2 were defined in the Kalimati Formation (Figure 1). Zone 8 at the basal part is characterized by variety of diatom assemblages and relatively low number of diatom valves, which reflect marsh or very shallow-water environments. Based on water-level indicators (ratio of planktonic diatoms to benthic diatoms and frequency of genus Aulacoseira), the water-level seems to have deepened gradually from zone 8 to zone 7, and retained deep condition from zone 6 to zone 4. Zones 6 to 4 are characterized by monodominance of Cyclotella: Cyclotella sp.1 is a characteristic in zone 5, and Cyclotella sp.2 is in zone 6 and zone 4. Number of their valves is very abundant. On the other hand, in zone 3 and zone 2, there are several dominant species and number of total diatom valves decreases. Furthermore, ratio of planktonic diatoms to benthic diatoms periodically rises and falls, which probably indicates water-level fluctuations. Especially in zone 2A, fluvial and marsh environments were expanded in marginal area of the lake, because relative abundance and number of Staurosira construens and Pseudostaurosira brevistriata, indicative species for marsh environment, become high. In zone 1, ratio of planktonic diatoms increases, which indicates the water-level rose again. But after zone 1, number of total diatom valves drastically reduces. It demonstrates that the lake was drained during a short period at about 12 ka. Comparison of environmental changes of the PaleoKathmandu Lake with δO record from a core MD900963 collected in the Indian Ocean (Bassinot et al. 1994) for the past 630 ka, shows that the water-level of the Paleo-Kathmandu Lake seems to have fallen at glacial age (MIS 14, 10, 8, 6, 2). We recognized following five major events expressed by high ratio of benthic diatoms, which indicate lowering of water-level. According to the time-scale derived from paleomagnetic study and AMS 14C dating of the core, the following stages correspond to marine isotope stages (MIS) from MIS 14 to MIS 2. Zone 5 ~ zone 4 = MIS 14 Early stage of zone 3 = MIS 10 Later stage of zone 3 = MIS 8 Later stage of zone 2C = MIS 6 Later stage of zone 2A = MIS 2 It is likely that these environmental changes of the PaleoKathmandu Lake mainly caused by fluctuations of Indian monsoon related to global climatic changes (glacial-interglacial cycle).