Documenting the evolution and transformation of natural history collection catalogues into digital online databases provides insight into the relevance and importance of processes, tools and decisions taken. This paper describes some important steps of the computerisation of the MNHN Fish Collection catalogue into the GICIM database and the onset of the collaboration with FishBase. The first specimens in the collection were catalogued in the second half of the 19th century. The initial estimate, in 1983, was 80,000 lots (a lot is a set of specimens caught in a same place in a same time and preserved in a same container) containing 1 million specimens. In 1994, about 82,000 lots were registered while most of the collection was computerised to become the GICIM database (Gestion Informatisée des Collections d’Ichtyologie du Muséum national d’Histoire naturelle / Computerised Management of the MNHN Ichthyological Collections). As of January 1st, 2022, 137,021 lots were registered containing 442,712 specimens. This early computerisation and database publication on the Web triggered a collaboration with FishBase. At the end of a first project in 1996-1997, georeferenced occurrence data for 78,000 lots were sent from GICIM to FishBase after standardisation and matching of the fields, in particular the taxon and country names. The subsequent development of the data entry interface and the structure improvement of GICIM made possible the management of additional information, and facilitated an analysis of the collection’s growth, and how the collection has been used over the past thirty years. This article highlights that specimens and their associated data are more valuable when they are placed in their historical perspective of acquisition facilitating quality control and interpretation. It also demonstrates that collections must keep the pace of the development of new technologies. Based on lessons learned during the 25 past years, the collaboration GICIM-FishBase will continue to push towards the exploration and application of new and more efficient methodologies.