Significant research has been undertaken over the past two decades to identify and delineate flood event and long-term runoff erosion during the Holocene period. Recent published work focused in the French Alps and Pyrenees has illustrated capacity to use lacustrine sedimentary infill, reconstituted form dated lake core samples, to discuss Holocene (and earlier) runoff and flood occurrences within alpine catchments. Lake sediments trap various paleolimnology proxies, which can be examined to paint a detalied petrography, in combination with age-depth modelling, gives us the possibility to estimate the amount of soil eorded within the catchment through time. In mountainous areas, such erosion fluxes are expected to be essentially driven bt rainfall and runoff (energy and climate divers). The work-package 5 from the TRAM research project proposes to study the sedimentary infill form Lake Arbu, a small high-attitude system of glacial origin located into the French Pyrenees, in order to investigate the potential of paleo-limnological dataset to provide clear runoff and erosion proxies potentially capable of long-term semi-quantitative catchment erosion analysis. Combining sediment accumulation rates with quantitative organic petrography,andhydrological dynamics (rainfall and runoff characteristics) longer-term sediment mass balance analysis can be completed. De Ploey's Cumulative Erosion Potential (CEP) presents one such simple mass balance model. Similar to the earlier USLE, CEP provides a mass balance analysis based on rainfall (and discharge) datasets, providing consideration of soil moisture capacity and erosion susceptibility (Es) for a selected catchment. Correlation analysis of sediment accumulation rates with mass balance expected deposition, keymass balance parameters (including CEP, Es and the RUSLE K), sediment land use and climate characteristics is a first step in identification of possible paleo-limnological contributing catchment erosion rate proxy definition. Where no single parameter has indicatively been found to effectively represent fine sediment deposition in the investigated mountain catchment sinks (lakes and peat), use of multiple regression and principal compoent/canonical correspondance analysis (PCA/CCA) is implemented to identiy key forcing parameters. The correlation and statistical analysis initially undertaken is the first step towards long-term sediment transport modelling to represent alpine catchment runoff and erosion over the Holocene time period, with the potential to consider future long-term or rapid climate change impacts on the mountain critical zone.