Accounting innovations, and especially the introduction of accruals accounting, are often portrayed as fundamental aspects of public sector reforms. An increasing body of literature has been studying the experiences of different countries and levels of government, often referring to Lueder's (1992) "contingency model" and its subsequent developments. This model highlights a number of social, political and administrative variables that are likely to affect government accounting innovations. It applies at a country-wide level and is mainly intented to describe and compare the enviroments within which accounting reforms take place, as well as the technical features of such reforms. Our paper, on the contrary, tries to expand the contingency model by applying it to individual organisations and by testing its explanatory power