For nursing home residents with advanced dementia, very little evidence is available to show clinical benefit from enteral tube feeding. Although no randomized clinical trials have been done, considerable evidence from studies of weaker design strongly suggest that tube feeding does not reduce the risks of death, aspiration pneumonia, pressure ulcers, other infections, or poor functional outcome. Nationally, however, utilization is high and highly variable. System-wide incentives favor use of tube feeding, and may influence substitute decision-makers, bedside clinicians, gastroenterologists, and administrators regardless of patient preferences or putative medical indications. Underlying the widespread use of this marginally effective therapy is a basic misunderstanding about malnutrition and about aspiration pneumonia. The face value of tube feeding is strong indeed. In addition to the general faith in intervention, the impulse to "do something" when things are going poorly, financial incentives favor tube feeding for gastroenterologists, hospitals, and nursing homes. The desire to avoid regulatory sanctions, bad publicity, and liability exposure creates a further incentive for nursing homes to provide tube feeding. Rational, evidence-based use of tube feeding in advanced dementia will depend fundamentally on improved education. Reimbursement schemes require significant modification to limit the irrational use of tube feeding. Nursing home regulations based more securely on scientific evidence would likely reduce nonbeneficial tube feeding, as would evidence-based tort reform. Quality improvement initiatives could create positive incentives. Realigning incentives in these ways could, we believe, improve the quality of care, quality of life, and safety of these vulnerable individuals, likely with reduced costs of care.