Farmland landscapes in developed countries have undergone major habitat changes over the past 60 years leading to the decline of many species. Of these, the little bustard, a medium-sized, long-lived, ground nesting bird, has declined by 95% in France over the last 35 years. Here we present the results of a 21 year survey of the nesting ecology of this elusive species, analyzing 157 breeding attempts, the largest data set ever collated for this species. Females had a strong preference for meadows for breeding, yet this habitat only represented 14% overall habitat. Alfalfa alone accounted for 50% of nest locations. However, apart from vegetation type, females did not show any other pattern of habitat selection (vegetation height, nest position within field, field under agri-environmental contract-AES-). In addition, the laying period was extremely extended, spanning almost 3 months. We did not detect any strong effect of crop, date or whether the nesting field was in AES or not, on clutch size, egg size and egg-laying date. However, there were long-term changes in breeding phenology (females breed earlier than 20 years ago), and selection of vegetation between years and within years. Hatching success was very low (about half of the broods were destroyed by farm work), and both fecundity and productivity per female were found to be approximately one third of the values expected for a stable population. Overall, nesting females of Little Bustards select meadows in regard to their availability, but do not show any particular preference within meadows' vegetation structure or height. We show that in such system, meadows act as ecological traps, and furthermore, because females do not appear selective, it is impossible to manage meadows in order to limit this trap. We finally analyze whether the land-sharing AESs can conserve this species in intensive arable systems and conclude that the land sharing may not be sustainable. We discuss our results in light of the alternative of land sparing, and suggest that this is probably a better fit for the conservation of large-bodied bird species given their ecological constraints (large home ranges, presence of semi-natural landscape components and freedom from human interference).