There are a growing body of theoretical work, wide anecdotal evidence, and a few large-scale empirical studies supporting the view that business firms quite rarely change their organizational structure, a phenomenon usually referred to in the literature as structural inertia. The present paper aims to analyze empirically the determinants of structural inertia and organizational change. As far as we know, this work constitutes the first attempt to directly address such issues through econometric estimates based on a large, longitudinal dataset at plant level. For this purpose, we consider changes of the organizational structure within a sample composed of 438 Italian manufacturing plants observed from 1975 to 1996. More precisely, we specify and test a duration model of the likelihood of an individual plant changing the number of hierarchical tiers after a spell r, provided that no change has occurred up to T. We also analyze the direction of change, distinguishing increases from decreases of the number of managerial layers. We consider a set of plant- and industry-specific explanatory variables that are expected to induce or oppose organizational change. The findings show that the adoption of advanced manufacturing technologies and new human-resources management practices favors organizational change. On the contrary, the presence of sunk costs and the extent of influence activities figure prominently in explaining structural inertia of business organizations.