Abstract Selected mode I fatigue data from five different types of fiber-reinforced, polymer–matrix composites tested in two round robins organized by the American Society for Testing and Materials subcommittee D30.06 and European Structural Integrity Society Technical Committee 4, respectively, are analyzed and discussed. The focus is on experimental scatter (in-laboratory and inter-laboratory) and on schemes for quantitative data analysis. It is shown that in spite of considerable scatter different composites can be distinguished and, under certain assumptions, a relative ranking be established. Further, effects from limited experimental measurement resolution are noted and implications for the test procedure and use of the test data in design of composite structures discussed. For comparative purposes, a rough ranking of different composites is feasible with test data generated within 24h per specimen in an industrial test environment.