Architecture of a software intensive system plays a key role in determining the achievement of architecturally significant requirements (ASR) of that system. If ASRs are wrong, incomplete, inaccurate, or lack details, then a software architecture based on these is also likely to contain errors. However, it has been a general observation, in the literature and practice, that it is quite hard to determine which are the architecturally significant requirements that an architecture team needs to consider while making the key design decisions. Whilst it is very clear that not all requirements are architecturally significant, it is quite different to identify those which are architecturally significant. This is partly due to a lack of an authoritative, evidence-based discourse characterising ASRs. To address this situation, we have proposed a framework for characterizing architecturally signifiant requirements based on empirical study using Grounded Theory. This work has been accepted in the special issue of IEEE Software on Twin Peaks of Requirements and Architecture (CharacterizingArchitecturallySignificant-ChenL-IEEE Software). The accepted article presents an evidence-based framework for systematically characterizing ASRs that can help practitioners to better understand and deal with ASRs. The framework can also facilitate systematic and structured discussion on the ASR research. We assert that this work can also enrich our understanding of requirements and architecture interactions, allowing the ‘twin peaks’ to move from aspiration to reality.