The paper uses ‘within-parent’ variation to study how changes in British parents’ health, marital status and financial resources affect receipt of help from their children. The analysis considers two measures of children's help (one enumerating specific activities and another reporting assistance with particular difficulties) and two measures of parents’ health: self-reported assessments of overall health and enumeration of difficulties with activities of daily living. It uses three longitudinal data sets from Britain: the British Household Panel Study, Understanding Society and the English Longitudinal Study of Ageing. The primary finding is that children's help is highly responsive to problems arising from a parent's health that limit a parent's mobility or their ability to live independently in a community. But it is not responsive to severe difficulties requiring daily care. The estimates of responsiveness that are based fully or partially on between-individual variation overstate the impact of parent's health on help.