As firms attempt to improve environmental performance at the plant-level, operations managers must decide the type of technological innovation that will be pursued. As a starting point, this paper considers the question: why are pollution prevention technologies not yet favored, relative to end-of-pipe pollution controls, in spite of the general acceptance that these technologies offer greater potential for improved competitiveness? Based on research in environmental management and technological innovation, this paper presents a conceptual framework that identifies two dimensions of technological innovation for environmental technologies: i) core components; and ii) system architecture. Modifications to the system architecture are further differentiated as either infrastructural or structural. Manufacturing competencies, organizational filters and internal coordination systems favor the pursuit of incremental innovation, or alternatively, innovation along only one dimension. Thus, end-of-pipe technologies, being the least innovative and most incremental in nature, tend to be adopted. Illustrative examples of environmental technologies are mapped onto this framework. Propositions for future research are offered, and implications for managerial action are summarized.