Edinburgh has 20-35 breeding pairs of Sparrowhawks. Their ecology is similar to those living in rural areas (See: Newton 1986). House Sparows were the most important Sparrowhawk prey: 27% of prey items, and 22% of biomass. Sparrowhawks fed on whatever prey, within a certain size range, was available. The mean nearest-neighbour distance was 1.3 ± 0.08 km. There was no indication that the composition of the population differed from that found in the rural areas. Adult female survival was 53%, yearling survival 41%. Pollutants were not particularly high. The average residency for a breeding female Sparrowhawk was 1.4 yrs, and 1.5 yrs for the male. 54% of all trapped females were new recruits to the breeding population. Range sizes varied with season, age and sex of individuals. The mean range size for breeding, adult males was about 400 ha. Range size could be affected by roosting behaviour. Some birds breeders roosted up to 2 km away from their nesting place. Range size of yearlings was directly related to the number of roost sites used, which increased as egg-date approaches. This was interpreted as prospecting for openings in the breeding population. 51 potential nesting places were identified, 39 were occupied in the 4 years. Some nesting places were preferred. On average, 24 nests produced eggs/yr, 3.94 eggs/nest were laid, 2.47 chicks hatched/nest, and 2.3 fledged/nest. Wet, cold weather in spring and summer decreased production. Production was affected by the turnover of breeding females. Other causes of variation, including human disturbance, could not be proven. The breeding behaviours of Sparrowhawks were observed and interpreted in terms of the possible reproductive strategies of the individuals. Breeding male and female Sparrowhawks spent more time at the nesting place as egg-date approached, and more time together in the morning and late afternoon. Mate guarding could not be proven. Displays increased as egg-date approached, and were more common in the morning. Courtship feeding increased as egg-date approached. Food presentations by the male were mostly made in the morning. Copulations showed similar seasonal and diurnal variation to the other behaviours. Copulation rates were high (150-0250/clutch). There was no concrete evidence that the male was buying copulations with food presentations. All behavioural rates were affected by breeding birds roosting at the nesting place together later in the season. 13% of all copulations were extra-pair, and occurred when an intruding female (which was paired with another male) solicited a territory holding male. Sperm storage tubules were identified in the reproductive tract of female Sparrowhawks. Chicks were produced which were not the offspring of the male with which the female was 'paired'.