For many years underground rock mechanics and in particular, roadway/tunnel roof stability has been underpinned by the often unchallenged assumption that roof strength (as defined by the UCS) and stiffness (E) are key stability controls. This has logically led to the proliferation of laboratory testing of rock specimens and the development of indirect geophysical methods to gain estimates of these two rock parameters. Furthermore, many design methods are significantly focussed on replicating rock mass behaviour through either intact or failed constitutive models. Demonstrably the strength and stiffness of the host rock material is commonly used as one of the key indicators of excavation roof stability and it finds either direct or indirect use in just about every rock mass rating system in use today. In more recent times there has been a common move to consider and apply (even if only conceptually at the current time) structural engineering type principles (eg, buckling) to coal mine roadway roof (and rib) stability. Similarly our knowledge of the in situ stress environment and its likely origins has improved significantly, largely through stress measurements and subsequent analysis. This paper combines knowledge in both of these fundamental areas through a deterministic model for roadway roof stability and in combination with field examples, reaches the almost certainly controversial conclusion that UCS and E are commonly irrelevant, albeit that the former may provide an indication of other relevant geotechnical parameters (eg, bedding cohesion). As with all hypotheses or rules, there are naturally exceptions and in this case, the most obvious is the tailgate of the longwall panel (with adjacent goaf). Due to the significant change in the strata loading environment of a longwall tailgate as compared to first workings for example, the stability equation materially changes so that UCS and E become critical controls. The point of the paper is to present a different perspective on a traditional mining problem and to challenge geotechnical professionals to keep thinking “outside of the square” in the never-ending endeavour to improve our understanding of the engineering problems we regularly face. Such an understanding impacts upon such issues as geotechnical data collection from borecore, support hardware requirements and design capabilities. Therefore making the assumption that our understanding is always fundamentally correct could in fact be limiting the development of new and improved engineering.