In comparison to studies in hydrology, progress in developing Geographical Information Systems (GISs) for catchment-based research in fluvial geomorphology appears to be slow due to problems relating to the digital representation of the channel perimeter, the non-incremental basis of river channel change and the extended timescales over which geomorphological change occurs. Acknowledging these constraints, critical system and parameter requirements are defined from which a GIS is implemented combining Laser-Scan mapping capabilities and macros with 'C' programs and ARC/INFO routines. Accurate data representation of linear and areal features is critical for deriving terrain and flow properties and an 'error' checking method for calculating upstream catchment areas is described. In general, system requirements are met with minimum data compromise but at cost to the system complexity and 'user-friendliness'. The potential for obtaining novel insights exists but requires careful attention to data quality standards.