Abstract This paper describes a real-time investigation of the total new product portfolios of nine product departments in the scientific instrument industry. It was found that within these portfolios only a few new products represented a major advance in functional performance. Most new products offered only incremental performance improvement, and many ‘new’ products were direct copies of competitors ' offerings. The scientific instrument innovation process is described in seven stages: recognition of need, invention of solution, prototype development, use in application, precommercial information dissemination, product engineering, and commercialization. In those cases where users were involved in the majority of the first five stages of the innovation process, the innovation was either a major functional improvement or a minor improvement that had not been initiated as a direct response to a competitor's product introduction. Direct copies and minor improvements that were competitive responses are notable for the absence of user involvement in the innovation process. Finally, the decision of a manufacturer to commercialize an innovation was unrelated to the level of user activity in the prior stages of the innovation process.