Routinely-collected statistics show considerable variation between local authorities in Great Britain, in the proportions of supported residents placed in nursing and residential care. This raises the question of whether this is due to variations in demand (the type of resident approaching authorities), supply (the level and type of provision available for local authorities to purchase), or policy (in terms of eligibility criteria or interpretations of need at field level). Data were used from a national longitudinal survey of individuals admitted to publicly-funded residential and nursing home care. Information was collected from 18 local authorities on a cohort of 2,544 local authority supported residents who had been admitted to residential and nursing home care. The paper examines the pattern of admissions, the characteristics of people admitted and the relationship between these characteristics and admissions to residential or nursing home care. Characteristics of the individual explained the placement of over 80 per cent of admissions. Supply factors were statistically significant but did not improve the explanatory power of the model. Survival among those admitted to a type of care that was not predicted by the model, suggested that some unmeasured aspects of prognosis may account for some of the residual variation in placements. Overall, the results indicate a reasonably high level of consistency between authorities in nursing home placement decisions. This suggests that either there is considerable variation in the types of individual approaching local authorities or, more likely, that some authorities are more successful in maintaining people for longer at home than others. In addition to maintaining people at home to a higher level of dependency, prevention of admission to residential care is likely to be associated with: interventions that address carer support, safety issues among people who are deaf, and motivation.