In the winter of 1984/1985 obseravtions have been carried out on (sea) birds during two weeks in December and two weeks in January at two gas platforms, owned by the Nederlandse Aaardolie Maatscbappij (NAM), in the North Sea, at approximately 120 km tot the northwest of Den Helder (North Holland). As so far surprisingly little is known, in spite of the work by Blake et al. 1984), about distribution and behaviour of birds at sea in the Dutch part of the Continental Shelf. So this paper pretends to summarize our first tentative and experimental steps to fill in the vast gaps in our current knowledge. A total of 11,820 minutes of observation has been accomplished, using the 180° scan method (for detials cf. Blake et al. 1984)and 132 counts of associated birds have been made. The results primarily indicate that, also for the more numerous species, strong daily and monthly fluctuations occur in the numbers of observed birds per unit time at a fixed point at sea. Apart from the influence of migratory behaviour and observation conditions, these fluctuation seem to be caused by, or at least correlated with environmental conditions, which in their turn again might influence migratory behaviour. Environmental conditions are not to be confused with observation conditions (e.g. poor visibility is an observation condition, whereas wind direction and -force is considered an environmental factor). It was far beyond the scope of this (experimental) try-out to intend to assess the influence of food on these fluctuations. Future work should, however, include such int'ntions. Furthermore, data have been gathered on behaviour and age-composition of several of the observed (sea)bird species in winter at sea. Pros and contras of platform observing at sea are discussed, as well as the methods used, which are compared with two other types of survey, used at sea, viz, ship-based and aircraft-based observations (cf. Blake et al. 19841. It is suggested that at least one of the assumptions of these two survey types, namely that one (or a few) trips are sufficient tot get a more or less accurate impression of the numbers of birds present in a certain area at a certain time of the year, should seriously be revaluated, considering the daily (and even hourly) fluctuations in bird numbers per unit time as found in this study. Nevertheless, future work in the Dutch part of the North Sea will have to depend heavily upon ship-based counts tot provide an overall picture of the entire area, independent of the influence of platforms on distribution of the birds. Platform observing, however, seems to be more appropriate to obtain more detailed information on migration movements (and the according fluctuations in numbers), feeding behaviour and exact plumage/age compositions of certain difficult species such as auks.