Safety-Critical Systems are becoming increasingly complex, with a trend towards automation and data exchange through emergent technologies under the banner of the Internet-of-Things. One such technology is the Digital Twin, a virtual entity aimed at replicating a real-world system ‘cradle-to-grave’, through a close bond. The lifeblood of that relationship is information, seen as enabling ‘sentient’ predictions in the Digital Twin, potentially mitigating unintended failures where its real-world twin is safety-critical. The defence industry is not immune with the emerging Internet-of-Military-Things, prompting consideration of what an assured Digital Twin of a safety-critical system might look like. Through consideration of the ‘state-of-the-art’, no substantive solutions were identified, leading to the development of a methodology to rationally support the formation of principles for Digital Twin assurance based on a refinement of the technique of Grounded Theory Analysis. When implemented and evaluated through objective reflection and peer review, the methodology resulted in a credible set of sector-independent principles for Digital Twin assurance of Safety- Critical Systems. This set is reported here.