The longest available bag record of Grey Partridges Perdix perdix in Great Britain (1793–1993) reveals a collapse of stocks after 1952 despite considerable annual variation. The annual fluctuations were attributable largely to annual variations in chick survival rate. The Game Conservancy Trust's National Game Census revealed that chick survival rates averaged 49% before the introduction of herbicides and 32% once their use became widespread. On a study area in Sussex, where spring density declined from around 21 pairs per km2 in 1968 to under four pairs per km2 in 1993, annual chick survival rates averaged 28% with no demonstrable trend. The annual over-winter “survival” rates in the area improved during 1968–1993, whereas brood production rates declined. Simulation modelling showed that a reduction in chick survival rate from 49% to 32% had little effect on spring stocks as long as nest predation was controlled but that stocks collapsed when nest predation control was relaxed. The effect of such a change in chick survival rate on population status was investigated by reference to 36 other studies in the literature. Amongst 20 studied populations which were stable, adjusting mean chick survival rates downwards produced demographic parameters characteristic of declining populations in all but two cases. Conversely, adjusting chick survival rates upwards for 16 declining populations made all but two stable. Diagnosing and remedying the causes of population change require a testable understanding of density-dependent factors and compensatory processes, best approached by a combination of monitoring, modelling and management.