This paper critically examines the concept of "localised knowledge spillovers" (LKS) and its role in the debate on the spatial boundaries of spillovers from private and public R&D. LKS are defined as knowledge externalities bounded in space, allowing firms near knowledge sources to innovate faster. They are central to econometric studies on the impact of academic and industrial R&D on firm innovation. However, the paper argues that LKS is a "black box" with ambiguous content, often used evocatively to signal interest in geography and innovation, but not to understand the mechanisms linking them. The paper highlights two main risks of LKS-based research: first, it may divert attention from the study of geographical distance's role in knowledge transmission, which remains controversial. Second, it may lead to naive policy implications, such as science parks and growth poles, which have had mixed results. The paper critiques the assumption that knowledge spillovers are a "stylised fact" and argues that they are not necessarily localised or tied to "tacitness." Instead, knowledge flows are complex, involving market mechanisms, labor markets, and appropriation strategies. The paper also challenges the concept of "tacitness" as an intrinsic property of knowledge, suggesting it is more about the messages exchanged within epistemic communities. It argues that knowledge can be shared across distances through various communication media, and that the property regime of knowledge is not limited to private or public goods but includes intermediate cases like club goods. The paper further examines the role of skilled labor mobility in knowledge diffusion, noting that while mobility can spread knowledge, it may also create tensions and congestion effects. It argues that the interpretation of labor mobility as a key mechanism for LKS is flawed, as it may not necessarily lead to knowledge spillovers. Finally, the paper suggests that rigorous research should consider not only the locational advantages of accessing academic research but also the diseconomies and the relationship between time and geography in spillovers. It emphasizes the need to explore both price and non-price mechanisms for knowledge transfer between universities and firms.This paper critically examines the concept of "localised knowledge spillovers" (LKS) and its role in the debate on the spatial boundaries of spillovers from private and public R&D. LKS are defined as knowledge externalities bounded in space, allowing firms near knowledge sources to innovate faster. They are central to econometric studies on the impact of academic and industrial R&D on firm innovation. However, the paper argues that LKS is a "black box" with ambiguous content, often used evocatively to signal interest in geography and innovation, but not to understand the mechanisms linking them. The paper highlights two main risks of LKS-based research: first, it may divert attention from the study of geographical distance's role in knowledge transmission, which remains controversial. Second, it may lead to naive policy implications, such as science parks and growth poles, which have had mixed results. The paper critiques the assumption that knowledge spillovers are a "stylised fact" and argues that they are not necessarily localised or tied to "tacitness." Instead, knowledge flows are complex, involving market mechanisms, labor markets, and appropriation strategies. The paper also challenges the concept of "tacitness" as an intrinsic property of knowledge, suggesting it is more about the messages exchanged within epistemic communities. It argues that knowledge can be shared across distances through various communication media, and that the property regime of knowledge is not limited to private or public goods but includes intermediate cases like club goods. The paper further examines the role of skilled labor mobility in knowledge diffusion, noting that while mobility can spread knowledge, it may also create tensions and congestion effects. It argues that the interpretation of labor mobility as a key mechanism for LKS is flawed, as it may not necessarily lead to knowledge spillovers. Finally, the paper suggests that rigorous research should consider not only the locational advantages of accessing academic research but also the diseconomies and the relationship between time and geography in spillovers. It emphasizes the need to explore both price and non-price mechanisms for knowledge transfer between universities and firms.