To investigate the role which standardised, patient related data play in indicated admission to a psychogeriatric nursing home, we compared the data of 120 indicated with those of 68 not-indicated patients. High scores on subscales 'helplessness' and 'inactivity' of the Dutch version of the Stockton Geriatric Rating Scale were the best predictors of indicated admission to a certain extent (21.6%). Furthermore a history of cardiac diseases and the state of being unmarried increased the probability of indicated admission to a small degree. Only a limited percentage of the variance in the decision of the indication committee could be explained. The limited availability of data which are collected in a standardised way, especially data concerning the caregiving system and the absence of valid and reliable instruments for the investigation of indicated admission are the most probable causes of this fact.