This letter notes that the work by S Korndorfer et al studying the long-term survival of patients with anorexia nervosa raises doubts about the excessive mortality rate usually associated with the disorder. Reference is made to an accompanying editorial that points out that the patients in the Kordorfer study had an increased risk of death by suicide, alcohol abuse, & other mental health problems but a lower than expected mortality rate from cardiovascular causes. Implications are discussed. (9 refs)