Existing regulatory definitions of time-periods over which evidence should be searched to demonstrate whether a fault is, or is not, active may be unconservative in slowly deforming continental regions where recurrence intervals of surface faulting are in excess of 100 000 years. Self-evidently, no definition of an active fault can be satisfactory when founded on some extrinsic time-period. Instead, it must be derived from the properties of the system itself, preferably through knowledge of fault-movement recurrence intervals, or alternatively and conservatively from an understanding of the time-period over which the present conditions of regional stress and strain (the ‘current tectonic regime’) have prevailed. Borrowing from volcanological terminology, out of the definition of the ‘active’ fault emerges an antithesis: the ‘extinct’ fault. Time-periods by which directly dated faults are designated as active or extinct have to be locally derived.