The Corn Bunting Emberiza calandra is one of the most severely declining farmland birds across Europe. In the UK, numbers fell by 86% between 1967 and 2008. Corn Buntings favour open landscapes, nest on or close to the ground, are often polygynous, double-brooded, and have a seed-based diet supplemented in summer by invertebrates. This study investigated the recent causes of decline in arable and mixed farmland in eastern Scotland, and sought to identify potential conservation solutions that could be delivered through agri-environment schemes (AES). Combining new data with analyses of existing long-term datasets, I investigated habitat associations during summer and winter, the timing and success of nesting attempts, and measured reproductive and population responses to AES. Corn Buntings declined almost to extinction in one study area where, over 20 years, the main recorded intensifications of farming were reduced weed abundance within crops and removal of boundaries to make bigger fields. Territory locations, late-summer occupancy and polygyny were all strongly associated with weedy fields. There were also positive associations with overhead wires and in early summer with winter barley and forage grasses. Late-summer occupancy was associated with spring-sown cereals, crops that are amongst the last to be harvested. Changes in habitat associations and to aspects of the mating system as the population declined and agriculture intensified are discussed. Intensive monitoring showed that Corn Buntings laid clutches from mid-May to mid-August, mostly in fields of forage grasses and autumn-sown cereals in early summer, and spring sown cereals in late summer. A preference for nesting in dense swards explained this seasonal variation. Breeding success in forage grasses was poor, due to high rates of nest loss during mowing. However, in experimental trials, nest success in fields with delayed mowing was fivefold that of control fields. With sufficient uptake through AES, delayed mowing could raise productivity to levels required to reverse population declines. In winter, cereal stubbles and AES unharvested crop patches were the main foraging habitats used. Unharvested crops with abundant cereal grain in their first winter of establishment were favoured. Population monitoring over seven years and 71 farms revealed increases on farms with AES targeted at Corn Buntings, no significant change on farms with general AES, and declines on control farms. In arable-dominated farmland, management that increased food availability reversed declines, but on mixed farmland where Corn Buntings nested in forage grasses, delayed mowing was essential for population increase. This study has already influenced the design of AES targeted at Corn Buntings in Scotland, and I make further recommendations for the species’ conservation and design of AES that are applicable to farmland throughout Britain and Europe.