Tsunamigenic events themselves leave other evidence in the geological record which can be used to infer the past occurrence of tsunamis. These differ according to the nature of the event but can be summarized as follows:

Earthquakes may produce coseismic subsidence or uplift, according to their geometry (see earthquake generated tsunamis). In certain parts of the world, notably Japan, Alaska and New Zealand, series of large, potentially tsunamigenic earthquakes are recorded by series of raised beaches which form terraced coastlines with the oldest terrace at the top. Dating of, for example, shellfish uplifted and killed by the earthquake then allows reconstruction of the earthquake history and hence also an estimate of tsunami recurrence interval. Sediment shaking by earthquakes also produces liquifaction in floodplain deposits and sediment flows offshore: these can also be identified and dated. Perhaps the most notable example of this is the work of Adams offshore from Cascadia (Adams, 1990): he showed that on thirteen occasions since 6850 years ago (the date of deposition of a volcanic ash bed at the base of the sequence), seismic shaking along the Cascadia subduction zone has produced simultaneous deposition of a type of sediment flow deposit known as a turbidite in submarine canyon systems several hundred kilometers apart. This result, together with the dating of the youngest turbidites as some 300 years old, is in remarkable agreement with the record of tsunami deposits on the adjacent coastline. Since such deep sea deposits are well preserved in many regions of the world, similar studies offer the potential for investigating the frequency of large, probably tsunamigenic earthquakes in areas where the conditions for tsunami deposit (sand sheet) preservation are not met.

Submarine sediment slides produce large, thick and laterally extensive deposits of sediment on the sea or ocean floor adjacent to the slopes on which they occur. In many cases, the sediment flows fast enough to produce a turbulent flow from which a giant version of a turbidite deposit (a sand, silt and mud sheet with highly distinctive characteristics) is formed. The turbidite formed from the sediment slide which produced the 1929 Grand Banks tsunami was in fact the first such deposit of this type to be recognized. In other cases, such as the Storegga slide, the sediment slide forms a distinctive submarine landslide. Many of these have now been mapped using sonar technology, as have the submarine landslide scars that they leave in their source areas, although the coverage of such mapping around the world is extremely patchy. Dating of the deposits without expensive sediment coring from research ships is also difficult and has only been carried out in a few locations.

Volcano lateral collapses also produce giant landslide deposits and leave lateral collapse scars. The latter can be tens of kilometers across and kilometers deep, most especially on oceanic island volcanoes (see Causes - Volcanic Eruptions). Careful mapping and radiometric dating also allows dating of these and hence estimation of the frequency of the giant tsunamis that they are believed to produce.

Asteroid impacts leave craters where they occur in shallow water on the continental shelf, as on land. However, deep ocean impacts, the most important from the perspective of tsunami generation, (see Causes - Impacts) expend most of their energy in producing a transient crater in the ocean itself and their signature in the geological record is more complex. The only near - impact deposit in the deep ocean that has been identified so far is that associated with the Eltanin impact in the southern Pacific, 2.16 million years ago: this consists of thick chaotic beds of sediment disturbed as the transient crater in the ocean collapsed in on itself.