Mental status changes were assessed and compared in 172 general surgical and orthopedic patients and 190 nonsurgical patients, all aged 55 and over, during a 10-month period. Assessments included a structured psychosocial questionnaire and standardized tests of cognition, affect, and function. The relationship of surgery, type of surgery, age, gender, and postoperative delirium to long-term postoperative decline was evaluated. Analyses of variance directly tested main effects pertaining to each of the five hypotheses and interactions of surgery with background variables. Hierarchical multiple regression analyses assessed the unique contributions of demographic and surgical variables to cognitive, affective, and functional change. None of the independent variables tested made a significant contribution to changes from baseline to long-term follow-up. The findings may be due to the physical and psychological health of this sample, and replication of this work in more impaired populations may be productive.