The available data are reviewed on ichthyofaunas from prehistoric sites along the Nile in Egypt and Sudanese Nubia. Former fishing practices are reconstructed using information derived from species spectra, reconstructed fish sizes, growth increment analysis and fishing implements. It is demonstrated that fishing was initially practised exclusively on the floodplain and that it was limited to a small number of shallow water taxa during Late Palaeolithic times. From the Epipalaeolithic onwards (ca 10000-8000 bp), fishing was also undertaken in the main Nile whereby the number of exploited species increased. Technological innovations allowing the exploitation of the deeper parts of the main river included nets and fish-hooks as well as improved vessels, permitting the capture of larger species from the open water. It is argued that fish must always have been a staple food because the animals seasonally occurring in large numbers on the floodplain were intensively exploited and because these fish could be easily dried for future consumption. Once the fishing grounds also included the main river, fishing was no longer restricted to the flood season, but could also be carried out when the Nile levels were low. Hence the role of fish in the resource scheduling also changed at the transition of Late Palaeolithic to Epipalaeolithic times.