In this review article we consider the crack growth resistance of micrometer and sub-micrometer sized samples from the fracture mechanics point of view. Standard fracture mechanics test procedures were developed for macro-scale samples, and reduction of the specimen dimensions by three to five orders of magnitude has severe consequences. This concerns the interpretation of results obtained by micro- and nano-mechanics, as well as the life time and failure prediction of micro- and nano-devices. We discuss the relevant fracture mechanics length scales and their relation to the material-specific structural lengths in order to conduct rigorous fracture mechanics experiments. To ensure general validity and applicability of evaluation concepts, these scaling considerations are detailed for ideally brittle, semi-brittle and micro ductile crack propagation, subject to both monotonic and cyclic loading. Special attention is devoted to the requirements for determining specimen size for various loading types to measure material characteristic crack propagation resistance at small scales. Finally, we discuss novel possibilities of micron and sub-micron fracture mechanics tests to improve the basic understanding of specific crack propagation processes. Display omitted •Fundamental concepts of fracture mechanics are revisited with respect to their applicability in miniaturized experiments•Stress state and notch sharpness in miniaturized samples need to be considered•For size independent fracture toughness values, sample sizes must obey relations to the microstructure and fracture process zone•Different contributions to material toughness can be separated by appropriate experimental design