The results of analyses of variability in seabird survey data aimed at improving predictions of seabird densities at sea are presented. Such predictions are important for informing conservation advice with respect to oil industry activities at sea, including potential or actual pollution incidents, and also for other users of the sea.