The study of life remote in space has strong parallels with the study of life remote in time. Both are dependent on decoding those historic phenomena called ‘fossils’, here taken to include biogenic traces of activity and waste products. There is the shared problem of data restoration from incomplete data sets; the importance of contextual analysis of potentially viable habitats; the centrality of cell theory; the need to reject the null hypothesis of an abiogenic origin for candidate cells via morphospace analysis; the need to demonstrate biology-like behaviour (e.g., association with biofilm-like structures; tendency to form clusters and ‘mats’; and a preference for certain substrates), and of metabolism-like behaviour (e.g., within the candidate cell wall; within surrounding ‘waste products’; evidence for syntrophy and metabolic cycles; and evidence for metabolic tiers). We combine these ideas into a robust protocol for demonstrating ancient or extra-terrestrial life, drawing examples from Earth's early geological record, in particular from the earliest known freshwater communities of the 1.0 Ga Torridonian of Scotland, from the 1.9 Ga Gunflint Chert of Canada, from the 3.4 Ga Strelley Pool sandstone of Australia, and from the 3.46 Ga Apex Chert also of Australia.