2010. The construction of a palaeodischarge time series for use in a study of fluvial system development of the Middle to Late Pleistocene Upper Thames.ABSTRACT: Recently many studies have attempted to model fluvial system development over a variety of geographical and temporal scales. It is generally recognised that one of the main problems, especially in studies over longer timescales (>100 ka), is the reconstruction of a robust palaeodischarge time series. Over such extended timescales discharge can only be reconstructed using proxy data, i.e. either field-based (sediment) palaeodischarge estimates or transformation of reconstructed palaeoclimate data series (e.g. ice core data), with only the latter method allowing the reconstruction of a continuous time series. In this study of the Upper Thames catchment, UK, we have developed a new palaeodischarge time series. A sea surface temperature record (ODP 980) from the North Atlantic (off the west coast of Ireland) is used as a proxy for precipitation across the Upper Thames catchment. A vegetation filter, based on pollen data, is then applied to this precipitation record in order to create a runoff model. Finally, this runoff model is transformed to a discharge model via the use of a climate change function which attempts to reflect probable changes in the frequency and magnitude of discharge events. Using our new palaeodischarge model, we present output from the FLUVER2 model of longitudinal profile development for the Middle to Late Pleistocene Upper Thames Valley. This model simulates the possible timing and magnitude of sediment aggradation/degradation events on the floodplain as well as the timing of floodplain abandonment due to tectonic uplift, resulting in terrace formation.