Age-specific survival and breeding (ASSAB) models were developed with data from 146 common buzzards (Buteo buteo) radio-tagged in southern Britain during 1990–1998, in a 120-km2 study area that had on average 25 egg-laying pairs. Survival checks were aided by philopatric behaviour and a maximum annual tag failure rate of 7%: minimum survival rates, that were estimated by assuming death of buzzards with lost tags, were close to maximum rates that were estimated using only the recorded deaths. First-year survival rate estimates for 35 buzzards fitted in 1990–1991 with 25–30-g backpack radios were 69–74% (minimum-maximum), close to the 61–71% for 16 buzzards with 12-g tail-mount radios; the backpacks transmitted for 2–4 yr. Overall survival rates were 66–73% in the first year, 91–97% in the second and 88–91% thereafter. Survival estimates from 288 recent British ring recoveries were lower in the first and second years, at 55% and 75%, but similar (88%) thereafter. Most deaths were from natural causes (40%) or interaction with artefacts (36%). ASSAB models, from radio-tracking and the observed 1.71 young clutch−1, predicted breeding by 16–21% of all the buzzards present in spring, or up to 25% with the minimum likely productivity of 1.4 young clutch−1 or 12% net emigration. Ringing data predicted breeding rates of 33–38%. The models were tested with density data from nest surveys and new radio-corrected-transect and truncation-mark-resighting estimates of buzzard numbers. Surveys in autumn and late winter estimated breeding rates of 21–25%. The high non-breeder density in spring, of three other buzzards for each paired bird with eggs, has important implications for understanding evolutionary fitness, predation and population ecology.