(1) Determine the relationship of geriatric assessment markers to 6-month postoperative mortality in elderly patients. (2) Create a clinical prediction rule using geriatric markers from preoperative assessment. Geriatric surgery patients have unique physiologic vulnerability requiring preoperative assessment beyond the traditional evaluation of older adults. The constellation of frailty, disability and comorbidity predict poor outcomes in elderly hospitalized patients. Prospectively, subjects > or =65 years undergoing a major operation requiring postoperative intensive care unit admission were enrolled. Preoperative geriatric assessments included: Mini-Cog Test (cognition), albumin, having fallen in the past 6-months, hematocrit, Katz Score (function), and Charlson Index (comorbidities). Outcome measures included 6-month mortality (primary) and postdischarge institutionalization (secondary). One hundred ten subjects (age 74 +/- 6 years) were studied. Six-month mortality was 15% (16/110). Preoperative markers related to 6-month mortality included: impaired cognition (P < 0.01), recent falls (P < 0.01), lower albumin (P < 0.01), greater anemia (P < 0.01), functional dependence (P < 0.01), and increased comorbidities (P < 0.01). Similar statistical relationships were found for all 6 markers and postdischarge institutionalization. Logistic regression identified any functional dependence (odds ratio 13.9) as the strongest predictor of 6-month mortality. Four or more markers in any one patient predicted 6-month mortality with a sensitivity of 81% (13/16) and specificity of 86% (81/94). Geriatric assessment markers for frailty, disability and comorbidity predict 6-month postoperative mortality and postdischarge institutionalization. The preoperative presence of > or =4 geriatric-specific markers has high sensitivity and specificity for 6-month mortality. Preoperative assessment using geriatric-specific markers is a substantial paradigm shift from the traditional preoperative evaluation of older adults.