Large numbers of seabirds may be killed from time to time by individual incidents of oil pollution, and throughout the year, especially in winter, dead seabirds, many of them oiled, are washed up on our shores. These dramatic events have given rise to a great deal of public concern about the effect of oil pollution on the wellbeing of seabird populations. It is important to consider this question from the point of view of population dynamics of seabirds so as to determine whether or not the observed mortality is substantial and additional in relation to the natural mortality. Such an approach requires detailed information on the distribution and numbers of seabirds at breeding colonies and at sea in their pre-breeding years, in association with their breeding activities and also in their `wintering' areas. While the data from breeding colonies provide censuses of breeding birds in defined geographical areas, movements and dispersal of breeding birds result in great uncertainty about their distribution and abundance while at sea. Since many oil polluting incidents, and much of the chronic oil pollution, affect birds while they are at sea, it is very difficult to asess the size and the provenance of the populations of various species that are actually at risk. Some evidence about the numbers of birds killed by oil and other causes can be obtained from both the beached bird survey, which are carried out monthly throughout the winter, and also from the recoveries of ringed birds. These sources of evidence give rather different results, but both are subject to difficulties of interpretation. Most seabirds are long-lived, with low mean annual adult mortality rates, and many of them do not breed until they are several years old. An attempt is made to relate the numbers of birds found dead, and the numbers oiled to the numbers that might be expected to die according to the measured rates of annual mortality. Again there are very considerable problems in attempting to relate these two sets of information. However, it seems that in western European waters the numbers killed by oil pollution is in tens of thousands per winter on average, while the number expected to die naturally is in hundreds of thousands per year. It is not known whether or not oil-induced mortality is additional to natural mortality. It is also pointed out that current environmental circumstances seem favourable and that the present resilience of populations may not persist if conditions change. Emphasis is placed on the very large numbers of pre-breeding birds and the need for information on the means by which they are recruited to breeding colonies. Recent monitoring of the numbers of breeding seabirds throughout Britain shows that most populations are increasing.