To help fulfil their responsibilities towards unpaid carers, service providers need some idea of the carer's situation and how many might require support. This paper argues that estimating the prevalence of unpaid care across service planning and budgeting cycles provides a better indication of the size and composition of the carer population than estimates at a point in time. The article presents prevalence rates of unpaid adult care from the British Household Panel Survey. It estimates the number of adults providing care at any time during the year for typical catchments or organisational settings, including social services and primary health care. It also provides related figures on carer turnover and changes in the carer population with an explanation of how they may be used and interpreted. As well as focusing on carers who are heavily involved in their caring activities, variations in the psychological well-being are assessed to provide and indication of unmet needs for support.