From 5 March to 1 July, 1999 and 2000, we made six shipboard surveys for finless porpoises (Neophocaena phocaenoides) in the Inland Sea, which is believed to be inhabited by one local population of the species. Each survey lasted 2-24 days and covered nearly the entire extension of the sea. In principle, we followed the method of Kasuya and Kureha (1979) and attempted to revisit as many ferry tracks surveyed by them as possible, but we used a greater number of observers for some selected surveys and added some new tracks. Records of strandings and incidental mortality of the species were also analyzed. Comparison of the two sets of survey data, which were about 22 years apart, did not show establishment of new habitat in the Inland Sea, but confirmed disappearance of some habitat previously occupied by the species. All the 18 tracks surveyed by both studies showed various degrees of density decline, and the decline was statistically significant for 12 of them. The sighting rate (number of finless porpoises sighted per 100 km survey) declined in both nearshore and offshore strata. Using the 18 tracks and combining all strata, we estimated that the current density was about 4% that of the late 1970s in the middle and eastern Inland Sea, while it was about 70% in the western Inland Sea. From these results, we conclude that the finless porpoise population in the Inland Sea has dramatically declined. This decline was consistent with the trends of strandings and mortalities in fishing gear. Entanglement in fishing nets is the only documented human-caused mortality for the population. However, the decline was probably a result of compound effects of various types of environmental degradation, including destruction of coastal habitats due to construction, chemical pollution, red tide and vessel traffic.