In 2009 the US Federal Aviation Administration (FAA) introduced a slow crack-growth approach for certifying composite and adhesively-bonded structures. This approach requires that the growth of a delamination or disbond is slow, stable and predictable under cyclic-fatigue loads. To predict growth in aircraft structures requires a methodology for translating laboratory crack-growth data to full-scale structures. Whilst this need not be a fracture-mechanics based approach, the present paper focuses on fracture-mechanics approaches since they have been widely adopted for this purpose for certifying aircraft structures. This approach uses the ‘similitude hypothesis’ combined with the concept of a crack-driving force (CDF) to link the results from laboratory tests to the cyclic-fatigue behaviour seen in full-scale aircraft tests. The present paper reveals that the range of the strain-energy release rates, ΔG, is not a valid crack-driving force. In contrast, in the present paper, a valid scheme is identified and proven to be appropriate.