A systematic review was conducted to determine the relationship between ratings of children's health-related quality of life (HRQoL) made by parents and children. This was investigated in relation to four questions: is agreement greater for some domains (e.g. physical HRQoL) than others?; do parents perceive illness to have a greater impact than their child?; how is agreement affected by child age, gender and illness status?; and is the relationship between proxy ratings affected by the method of data collection? Fourteen studies were identified. Consistent with previous research, there was greater agreement for observable functioning (e.g. physical HRQoL), and less for non-observable functioning (e.g. emotional or social HRQoL). Three studies assessed whether parents perceive the illness to have a greater impact than their child, but no clear conclusions could be drawn given differences in measures used. Agreement is better between parents and chronically sick children compared with parents and their healthy children, but no effects were found for age or gender. All of these results may be dependent on the specific measure of HRQoL employed. There remain strong arguments for obtaining information from both parents and children whenever possible.