The purpose of the British Museum's Ashurbanipal Library Project is to investigate the content of the significant tablet collection that this Assyrian king assembled for his royal library. The initial project is focused on the Babylonian texts in order to establish the compositions involved and their relation to the rest of the Kouyunjik Collection and to the collecting activities of Ashurbanipal (668–627 BC).

The examination of the Babylonian texts of Ashurbanipal's library is a difficult task. Whoever is familiar with the Nineveh texts knows that the tablets were originally stored in four different buildings (see Fig. 1): in the South-West Palace, in the North Palace, and in the vicinity of the temples of Ištar and Nabû, with some additional find spots on and off the mound Kouyunjik. It is the tablet collection of the South-West Palace that formed the library of Ashurbanipal, but the excavation reports of Nineveh very seldom refer to the places where the tablets were found. To reconstruct the different libraries and archives is a very time-consuming task and beyond the possibilities of the six-month timetable for this project. Therefore, for the time being, I decided to consider the Babylonian literary tablets and all legal documents written during the reign of Ashurbanipal and his predecessors as coming from one place, namely Ashurbanipal's library or libraries at Nineveh.

While surveying the approximate figure of 26,000 tablets and fragments that the British excavators unearthed in Nineveh I entered the genre and content of the Babylonian texts in a database, together with a short description of the fragments, e.g. shape, colour, number of columns, lines and dividing lines. This database includes information on about 4290 tablets and fragments, of which 610 have already been rejoined to other fragments. Therefore, until now, the total number of Babylonian texts and fragments excavated in Nineveh is about 3680 — or in other words about one-seventh of all of the British Museum's Nineveh collection. The database I created also serves as a basis for collecting all texts of the same kind in order to identify joining fragments.