The OSPAR Revised Eutrophication Monitoring Programme (ETG 05/3/Info.1-E)requests that nutrient “monitoring should include sufficient samples to confirm that themaximum winter nutrient concentration has been determined”, while para. 7 of theTerms of Reference for the preparation of guidance on the spatial and temporalresolution of monitoring for nutrients and eutrophication effects (ICG 003) states that“there are at least nine different water types covered by the OSPAR MaritimeArea”…”guidance must, therefore, be at least complex enough to cover each type withsufficient clarity to guide contracting parties in their evaluation of the temporal andspatial coverage required to adequately assess the relevant water body”.This document summarises the national reports submitted to the OSPAR IntersessionalCorrespondence Group on Eutrophication Monitoring, and highlights commonproblems faced in the monitoring of (primarily) inorganic nutrients and chlorophyll. Inaddition, it presents tests of different approaches to solving the spatial and temporalsampling problems associated with delivering marine eutrophication data.Based on tests of model data, monthly sampling appears adequate to give a goodestimate of annual mean concentrations. Buoy data suggests that this would not besufficient where there is tidal variability. It was not possible to determine maximumconcentrations through a practical ship sampling scheme, or by using extreme valuestatistics.The optimum sampling programme to observe rapid events is likely to be a combinationof ferrybox systems, which appear to be reliable and give both good spatial andtemporal coverage, combined with buoy observations. To ensure data of sufficientquality, these must be controlled against conventional research vessel observations.Research vessels still have a role in seasonal mapping, and in providing data ofsufficient quality for trend analysis from a large area. This is likely to remain so, at leastuntil technologies such as gliders and optical nutrient sensors become widely availableand capable of delivering reliable, high quality data.