This paper presents an agent-based model of industry evolution in which technological and demand conditions contribute to determine both the emergence and performance of spinoff firms. We assume that two main factors characterize the emergence of spinoffs: first, spinoffs have at least some degree of product differentiation with respect to their parent; second, spinoffs share at least some elements with their parent, in the form of innovative capabilities transmitted to employees and research teams. The model is able to generate the stylized facts uncovered by empirical research, suggesting however that they might be the outcome of different processes in different sectoral contexts.