This study of Sparrowhawks was conducted in the southern district of the city of Rotterdam. The study plot encompassed 29 km² and consisted mostly of high density residential areas with a scattering of parks, public gardens, industrial development and sporting fields. The major part of this (sub)urban region was systematically surveyed for Sparrowhawks by recording visual observations, signs of its presence (plucking sites, droppings, pellets, moulted feathers) and nests. Prey remains were collected from May through August 2002 (considered as prey captured during the breeding season) and from September through April in 2000-03 (outside the breeding season). During the breeding season of 2002 twelve territories were located in Rotterdam-Zuid (0.41 territories/100 ha). Presumably this is an underestimate as private gardens and some quarters were not, or not completely, covered. Recorded breeding was restricted to parks and public gardens in the vicinity of residential areas with a high prey density. Mean nearest neighbour distance was 908 m (SD=263, median 700 m, range 620-1820 m, n=10 nests). Nesting trees were Pinus nigra (3), Salix sp. (1), Populus sp, (1), Betula pubescens (1), oak Quercus sp. (1), Ulmus sp. (1), Amelanchier lamarckii 91) and Acer campestre (1). The high frequency with which deciduous trees were used for nesting reflects the scarcity of coniferous trees in the study area. Sparrowhawks were recorded in all types of habitat; parks and gardens were used as resting and nesting sites whereas residential and industrial areas were exploited during hunting forays. In 10 out of 12 territories breeding was recorded (nest found), of which 9 pairs succeeded in raising at least one fledgling (1x ? number of chicks, 2x 1 chick, 1x 3, 4x 4 and 1x 6 chicks; recorded from the ground). One pair apparently did not lay eggs. Non-breeding was recorded in two territories (no nest found), involving a solitary bird (probably female) and an adult pair. In the latter case, it is possible that a failed breeding attempt was overlooked. The presence of floaters was indirectly substantiated when a first-year nesting female was found dead during the egg-laying stage; this bird was immediately replaced by another first-year female, using the same nest and successfully raising one chick with her first-year partner. Six pairs consisted of an adult male and an adult female, one pair of an adult male and a first-year female and one pair of both sexes in first-year plumage. During the breeding season, 22 bird species and 1 mammal species were recorded as prey (Table 1, N=329). Mean prey mass was 39.3 g. Four species dominated the diet, i.e. Passer domesticus, Parus major, Sturnus vulgaris and Parus caeruleus (together 82.6% in numbers and 63.9% in biomass). The proportion of juveniles among House Sparrows increased in the course of the breeding season: 0% in April, 25% in May, 47% in June and 53% in July. Of Starlings, 91% of the captures related to juveniles. Outside the breeding season, 242 pluckings consisting of 21 bird species and 1 mammal species were collected. Mean prey mass was then 112.4 g. Starlings were particularly important as prey (40.9% in numbers, 25.5% in biomass), mainly because of the presence of roosts in the area where several Sparrowhawks hunted simultaneously. Together with House Sparrow, pigeons, doves and thrushes, they constituted 88.2% of all prey items (95.6% of biomass) outside the breeding season. The city life of Sparrowhawks was also reflected in the capture of several non-native escapes, such as parakeets and a lovebird. In Rotterdam, breeding Sparrowhawks showed a high tolerance to human activities, with one female even feeding nestlings in full view of unaware pedestrians. Several nests were situated within few metres of footpath, bicycle path or road. Human disturbance was not recorded, nor predation by Magpies Pica pica, Carrion Crow Corvus corone or Jay Garrulus glandarius (partial egg predation could not be detected because nesting trees were not climbed). Northern Goshawks Accipiter gentilis are still not breeding in the city of Rotterdam, and Tawny Owls Strix aluco are rather scarce. It is therefore argued that presently the city of Rotterdam constitutes a safe breeding haven with a high prey density. Food competition with Eurasian Kestrel Falco tinnunculus, Eurasian Hobby F. subbuteo. Tawny Owl and domestic cats Felis catus is unlikely, either because of scarcity of potential competitors, different timing of the breeding cycles or different diets. Main threats are habitat changes, especially in the wake of renovation of old city districts and the disappearance of small parks through development, negatively impacting nesting sites and abundance of prey species.