Background: To investigate the characteristics of patients in whom artificial nutrition and hydration (ANH) is forgone, duration of survival after the decision and factors that are associated with duration of survival.Methods: Observational study based on written questionnaires in 32 Dutch nursing homes. Of 178 nursing home patients with dementia, their treating nursing home physician (NHP) filled in a questionnaire directly after the decision was made to forgo ANH. The maximum follow-up was 6 weeks. Cox proportional hazards analysis was used to determine predictors of survival.Results: Decisions to forgo ANH in Dutch nursing homes were made most often in patients with severe dementia who also had an acute illness. More than half the patients (59%) died within 1 week after the decision. Patients with dyspnea and/or apathy were more likely to die during follow-up than patients without these symptoms. Patients who were considered more severely ill by the NHP were more likely to die than those who were considered less severely ill. The presence of restlessness indicated a higher chance of survival.Conclusions: The clinical judgment of the NHP of the severity of illness appeared to be a strong predictor of patient survival. NHPs should not rely solely on their clinical judgment concerning survival, but they should also consider the presence or absence of dyspnea, apathy and restlessness.