Four hundred consecutive hip fractures were studied prospectively. Two hundred forty-seven patients were classified as unhealthy (poor cardiac status, pneumonia, cancer history, bowel obstruction history, malnutrition, dehydration, stroke history, renal failure history, cirrhosis). Twenty-two percent of this unhealthy group died, while only 6% of the remaining healthy group died. Death rates varied with admission activity level and mental status but not when patient health status was factored out. After factoring out health status, age was associated with higher death rates only in patients older than age 85. Confusion, a change of mental status in the hospital, occurred in 25% of patients. Confusion was associated with a medical complication in 94% of cases, was the presenting symptom of a medical complication in 79% of cases, and was associated with a 39% death rate. Major medical complications occurred in 9% of the healthy group (29% of them died) and 21% of the unhealthy group (64% of them died). Major medical complications in unhealthy, shut-in patients were associated with an 80% death rate. Vigorous urinary tract monitoring and early treatment of bacteriuria decreased death rate. Postfracture malnutrition was associated with higher complication rates. Hip surgery performed within 72 hours on patients with acute medical illnesses in addition to their fracture was associated with a higher death rate. Whether a patient walked postfracture seemed not to be correlated with the death rate. Patients who were not walking prefracture but treated by internal fixation had a 34% failure rate.