The analyses were based on the comparison of three wood-dwelling, medium sized raptor species with 366 nest-years in Buzzards Buteo buteo, 110 nest-years in Goshawks Accipiter gentilis and 34 nest-years in Honey Buzzards Pernis apivorus. Field data from 52 different nesting-habitat plots of Buzzards, 25 plots of Goshawks and 28 plots of Honey Buzzards were evaluated by discriminantanalysis and revealed differences in nest-habitat selection between these species. Nesting-habitat features (measured in these plots) which were found to be associated with habitat selection in a principal-component-analysis, were also related to reproductive success with some differences between the species. These differences were similar to the different habitat selection found in the discriminant- analysis. The gradient (from poor to good) in habitat selection is slightly different between the species considered. Goshawk and Buzzard were found to be more closely related in habitat use than either were to the Honey Buzzard. None of the species used their nesting-territories at random from year to year. Therefore, certain places had nests more often than expected by chance at the population levels found and others less often. (5) This use was related to nest-success. On average, the most often used territories showed the greatest success in all species. From both parts of my analyses the same nesting- habitat features were associated with good breeding success: habitat-parameters which gave good correlations to breeding success were found to be overrepresented in the more frequently used territories.