This paper develops a knowledge-based model of information systems (IS) integration in acquisition-based growth programs. Previous research has found important differences in the acquirers’ abilities for acquisition IS integration, and that these differences play key roles in explaining the economic benefits of acquisition-based growth. A knowledge-based view of acquisition IS integration contributes explanation to how differences in abilities arise. Our knowledge-based model is developed by an analytic induction approach, comprising both deductive and inductive analysis of the industry group Trelleborg. The analyses yield five propositions that explain the building of knowledge for acquisition IS integration by the mechanisms of routine refinement, superstitious learning, expertise building, sub-activity refinement and related expertise building. Together, the five propositions form a knowledge-based model of acquisition IS integration. The model suggests that ‘serial acquirers’ form the knowledge necessary for attending to an IS integration challenge over a series of heterogeneous acquisitions. Differences in acquirers’ abilities for acquisition IS integration that depend on knowledge variations are therefore persistent and hard to overcome for the inexperienced acquirer.