During January through March 1995, the activity of Common Buzzards was quantified from a car during 6931 observation minutes on 24 days. The study site Witterdiep (S of Assen, 52°58’N, 6°31’E) consisted mainly of grassland with fences or ditches in between fields, some woodlots and hedgerows, an artificial embankment and no human settlements (see Photos I and 2). Common Buzzards do not breed in this area, but the nearest nests were less than 1 km away. The study area held two winter territories: the northern territory (33.6 ha) was claimed by an adult male and a first-year female, the southern territory (50.7 ha) by an adult female (Fig. 1). Up to three floaters visited the area, but their visits were normally of short duration because of agonistic behaviour of the territorial females. At least two floaters were eventually found weakened and unable to fly; these birds probably died in late February or early March (Photo 3). A third floater involved an adult male which showed interest in the first-year female of the northern territory. Territorial conflicts were infrequent, and decreased in frequency during January through March (Table 3). On the other hand, agonistic behaviour of Carrion Crows against the Buzzards increased through time: 0.06 conflicts per 100 observation minutes in January (n=l), 0.26/100 min in February (n=7) and 0.84/100 min in March (n=21). As a consequence, Common Buzzards increasingly started to use sitting posts in the shelter of hedgerows and woodlots. Most time was spent wailing on poles (42% of total time) or in trees (39%), and less often on the ground or on the embankment (Table 1). The duration of each stay was highest in trees: on average 15.8 min. Much briefer sessions were spent on poles or on the ground, normally shorter than 5 min. The long average stay in trees is, however, an artefact of several extended periods of shelter-seeking during severe weather. The median duration of stays did not differ much between poles, trees, ground and embankment (Table 1). The birds frequently changed position, especially so in March (Table 2). Normally, periods of active hunting were alternated with prolonged periods of inactive behaviour. Stays of >30 minutes were usually spent in the shelter of trees and stays of >1 hour were witnessed in only 12 out of 1071 observations. Overall, only 3.2% of the observation time was spent flying. Flights between sitting posts had an average duration of 9.1 sec (n=342, SD=10.5, range 1-56, median 6). Longer flights involved soaring and heading for roosts outside the study area. Soaring only took 0.3% of the time observed in January and February, slightly increasing to 1.5% in March. The latter value is a minimum, because soaring birds were thrice lost from view and could not be accurately timed. Soaring by Common Buzzards from the study area was witnessed on 1 out 6 days in January (low), on 1 out of 8 days in February (low) and on 5 out of 10 days in March (also high). Presumed local breeding birds outside the confines of the study area soared much more frequently, i.e resp. on 1 (out of 6), on 4 (out of 8) and on 8 (out of 10) days in resp. January, February and March. Hovering was observed on two days with stormy weather, with clocked durations of 256, 211, 190 and 30 sec. Hunting success and prey selection were difficult to observe accurately. Presumably, the birds mainly preyed upon common voles Microtus arvalis, but hunting success could not be determined. Worm-eating was repeatedly observed, especially by the first-year female while walking on the embankment. On 28 January, she took 113 earthworms Lumbricus terrestris in 15 minutes, on 31 January 204 earthworms in 24 minutes and on 2 February 33 earthworms in 16 minutes.