A standardised behaviour rating scale, the Shortened Stockton Rating Scale, was completed by care staff for each of 903 residents in old people's homes. In an attempt to validate the four subscales suggested by the authors of the published scale, the data were subjected to factor analysis. The results indicate the existence of three factors reflecting physical, mental, and social impairment. They offer only partial support for the published subscales and the need for such scales to be validated on the intended target population is discussed.