Palaeoceanography relies on the assumption that parameters measureable in sediment cores correlate reliably to hydrographic parameters. Areas of dynamic watermass mixing, which generally have steep hydrographic gradients, therefore provide both a tempting target (large spatial and temporal differences in temperature and salinity) and a significant challenge (high flow velocities and high "noise") for palaeoceanographers. Here, we investigate the ability of parameters measured on core tops to be used as the basis for simulating regional hydrography within one globally important area of watermass mixing, the Gulf of Cadiz. Using grainsize, sediment composition, benthic foraminiferal assemblage and benthic foraminiferal stable isotope analysis it is possible to build an effective framework for qualitative constraint of the position of the modern Mediterranean Outflow Water (MW) Plume. We propose that the Gulf of Cadiz slope can be sub-divided into 8 hydrographic zones reflecting these parameters: Proximal MW, Core MW1 (steep slopes), Core MW1 (gentle slopes), Distal MW1, Distal MW2, Lower Limit of MW plume, North Atlantic Deep Water and Ambient Atlantic Water. We anticipate that compilation of time-slice data will reveal hydrographic sub-divisions of the slope for the past in a similar manner, improving our understanding of past changes in the size and position of the MW plume, but more work needs to be done before a secure protocol for quantitative reconstruction can be created. Stable isotope analysis alone is insufficient for the task, with delta C-13 behaving in a non-conservative manner and delta O-18(calcite) acting in an ambiguous way due to the competing influences of delta O-18(water) and temperature. It is likely that if a single parameter that unambiguously determines the difference between Mediterranean Water and Atlantic Water can be identified, combination with delta O-18(calcite) will allow extension of our qualitative analysis into a quantitative means of palaeohydrographic reconstruction for this region. Similar complexity to that found in this study would be expected in any mixing region, and in particular on the majority of sediment drifts. We recommend that core top surveys similar to this study are performed on these sediment drift systems before any form of quantitative palaeohydrology is attempted. (C) 2011 Elsevier B.V. All rights reserved.