The Life Skills Profile (LSP) was developed as a measure of function and disability in schizophrenia. We now report a number of studies assessing its psychometric properties, principally when completed by case workers, residential carers and parents. We demonstrate high test-retest reliability for all 3 rater groups. High interrater reliability was demonstrated between raters of the same background (such as parent vs parent) and when co-ratings by caseworkers and residential carers were examined. By contrast, parents rated differently to case workers and residential carers, a finding that could reflect different raters "knowing" subjects in different ways, or the effects of response biases. To pursue the latter possibility, the influences of rater "burden of care" and "optimism" were examined. A concurrent validity study is reported, offering additional support for the validity of the measure.