• There is a clear need to predict the occurrence of marine mammals in order to minimise the possible harmful impact of military sonar activities, some of which have recently received extensive public media exposure. • No military or civilian method currently exists to predict the possibility of encountering marine mammals. • The proposed work will exploit components of existing NERC-funded research within the core science programmes of SAMS and SMRU to develop a predictive tool that will link marine mammal occurrence to classical oceanographic observables. • Pre-existing data on marine mammal aggregations lead us to believe that the proposed method has a high probability of success. • The main study area will be the open seas to the north and west of the Hebrides. • Existing NERC-funded SAMS cruises in this area will collect oceanographic data, supplementing remotely sensed imagery and the extensive SAMS archive of time-series from this area. • A key element in achieving the proposal objective (and in furthering NERC science objectives) will be the recruitment of SMRU observers and equipment to SAMS cruise complements so that marine mammal sightings may be linked directly to the oceanographic research programme. • Additional SMRU deployments on board vessels of opportunity will increase the density of the observational programme. • The suitability of SOSUS acoustic data as an indicator of marine mammal presence will be investigated. • The datasets will be analysed using a variety of statistical techniques to yield a practical relationship between observables (local oceanographic conditions, season, location) and species abundance. • The value of the relationship as an operational tool will be tested in a variety of scenarios. • The work is expected to run from the summer of 2002 to the summer of 2005.