This talk is a nostalgic journey through the first 25 years of groundwater modelling in Australia, roughly the years 1970 to 1994 (when PEST was born). The early years were before MODFLOW, before Excel, before SURFER, before Windows, and long before GIS. Back then, every modelling activity was an innovation. Our perspective is coloured by personal experience in New South Wales (Australia) when the authors were working together in the State Government Water Conservation and Irrigation Commission. The game-changer was acquisition of a Digital PDP8/e in 1972 with 8 kb of memory, which supported only Assembler and FOCAL programming languages, followed by BASIC and eventually FORTRAN. I/O was via hardware toggles, a teletypewriter keyboard (10 cps), and paper tape which doubled as the initial storage device, followed by serial cassette tape, 8 inch floppy discs and eventually large hard discs with 5 Mb capacity. The focus was on analytical models for the first decade until distribution of PLASM (from USA) in the mid 1970s, AQUIFEM in the early 1980s, and MODFLOW in the mid 1980s, supplemented by bespoke code. Our story ends with the creation of PEST, a great Australian invention. In keeping with the theme of this conference, we witnessed groundwater changing from obscurity to prominence over a quarter of a century, and analytical and numerical techniques developed by us and others in those early years were truly “emerging”.