Academics are under increasing pressure to demonstrate the impact of their research with external actors. Some national research assessment systems have mandated academics to document their impact on non-academic actors, and linked research funding to assessments of these impacts. Although there has been considerable debate around the design of these systems, little is known about how academics perceive the value of impact against more conventional academic outputs, such as publications. Using multisource data, including a large-scale survey of UK business and management academics, this paper explores the individual and institutional factors that explain an individual’s preference for impact versus publication. The results show that academics display a preference for impact over publications, even when that impact is not associated with requirements of the assessment system in terms of rigour of the underpinning research. The preference for impact over publications is heightened by organization tenure, non-academic work experience, intrinsic career motivations and research-intensive contexts, while it is weakened by academic influence, extrinsic career motives and academic rank. We explore the implications of these findings for the design of research assessment systems and academics’ reactions to them.