The Mw 7.8 14 November 2016 Kaikoura Earthquake generated many thousands of landslides. In the days and weeks after it our team of scientists started mapping the landslides it generated. Initially this was done using low-resolution satellite imagery captured immediately after the earthquake. Between November 2016 and January 2017, an aerial photograph survey using a fixed wing aircraft was carried out. This survey captured a series of high-resolution photographs of the main area affected by landslides. We used these images and those from a previous aerial survey carried out in 2015, before the earthquake, to map more of the landslides. The Version 1.0 landslide inventory contains 10,195 hand-digitised landslides and was completed in August 2017. This version contained mainly the larger landslides (with source areas >10,000 m^2) triggered by the earthquake. The Version 2.0 inventory contains 29,557 hand-digitised landslides and was completed in August 2019. Version 2.0 contains all landslides from Version 1.0 – which were modified where necessary – plus additional landslides (with source areas mainly <10,000 m^2) mapped from the high-resolution aerial photographs. Whilst Version 2.0 contains significantly more landslides than the Version 1.0 inventory, it does not contain all the smaller landslides (with source areas typically <500 m^2) that were triggered by the earthquake, as differentiating such small landslides from apparent “noise” e.g., shadows, vegetation changes and other landslides was difficult and time-consuming. Both Version 1.0 and 2.0 inventories published here are points that represent the centroids of the mapped landslide source areas thought to have been triggered by this earthquake. The version 1.0 data is in CSV format, and the Version 2.0 data is in both CSV format and in GIS Shapefile format. The data are projected in the NZGD_2000_New_Zealand_Transverse_Mercator Coordinate System. These data sets can be used by others to explore what factors contribute to the occurrence of a landslide during an earthquake, as well as investigating how landscapes change as a result of large earthquakes. Our plan is to eventually publish a Version 3.0 inventory, which will contain the landslide source and debris polygons, along with other landslide attributes such as volume.