The reconstruction and tracing of transport paths in glaciated (and other) environments have a long tradition in the Earth Sciences. We here present a dataset of clast shape samples from a worldwide selection of glaciated mountain environments in order to assess the reliability of this approach overall and the role of lithology on the performance of clast shape measurements in particular. Our findings demonstrate that the widely-used RA-C co-variance approach is applicable to 63% of the 19 catchments investigated, while the alternative RWR-C approach is more widely applicable to 75% of these catchments. A systematic assessment of mixing of lithologies at the catchment scale demonstrates that such mixing leads to pronounced overlaps between different control envelopes that had previously been separated, thereby removing the discriminatory power of the method. Mixing of similar lithologies between different catchments shows an even more extreme loss of discriminatory power, which strongly suggests that lithology plays a primary role in determining clast shape, and that catchment-specific processes are superimposed. Systematic analysis of the dataset also shows that nearly all catchments (apart from two) can be grouped into two types. Type I relates to sites in lesser mountain ranges and is characterised by dominantly blocky forms in the subglacial realm, highlighting significant reworking processes. Type II sites are dominantly in high-mountain environments and characterised by a high similarity between subglacial and fluvial control envelopes. This indicates that, although reworking may be pronounced, it is not necessarily effective enough to remove the platy shape that most likely results from extraglacial and supraglacial inputs. Our study highlights the potential of clast shape analysis as a tool that allows generic processes to be identified between catchments, thereby enabling an understanding of debris cascades in glaciated mountain environments. We finish with recommendations for ensuring that future clast shape studies are robust, reproducible and comparable between different sites.