Patients in long-term care facilities are often moved from one living area to another within an institution. While the staff often consider such moves as part of the daily work load, even intrainstitutional relocation has been observed to have negative consequences for the elderly patient. If staff could identify those patients who might experience difficulties with a relocation, they could deal more intensively with these high-risk patients before, during, and after the move, hopefully avoiding a crisis situation for the patient. Staff and service professionals, however, may not be particularly accurate in assigning risk status. This study was undertaken when it was decided that a large number of nursing home care patients within a Veterans Administration domiciliary setting would be relocated upon completion of a renovation program. It attempted to identify high-risk relocation patients and to examine the extent to which staff ratings of patient risk status could identify such patients. Male nursing home residents (N=116) requiring minimum or moderate care were rated by 16 registered nurses, 3 licensed practical nurses, and 6 nursing aides. Ratings were averaged within level of training. The means for the three levels were significantly different and within each level there were significant differences among individual staff members. The findings suggest that the staff often misperceive the elderly and question the reliability of staff ratings as a means of identifying patients at risk from an impending relocation. It would appear that additional techniques are needed to overcome the lack of consensus among staff members. (NB)