Critically ill patients are more susceptible to malnutrition due to their severe illness. Moreover, elderly patients who are critically ill lack specific nutrition recommendations, with nutritional care in the intensive care units (ICUs) deplorable for the elderly. This study aims to investigate nutrition treatment and its correlation to mortality in elderly patients who are critically ill in intensive care units. A multiple-center prospective cohort study was conducted in China from 128 intensive care units (ICUs). A total of 1,238 elderly patients were included in the study from 26 April 2017. We analyzed the nutrition characteristics of elderly patients who are critically ill, including initiated timing, route, ways of enteral nutrition (EN), and feeding complications, including the adverse aspects of feeding, acute gastrointestinal injury (AGI), and feeding interruption. Multivariate logistic regression analysis was used to screen out the impact of nutrition treatment on a 28-day survival prognosis of elderly patients in the ICU. A total of 1,238 patients with a median age of 76 (IQR 70-83) were enrolled in the study. The Sequential Organ Failure (SOFA) median score was 7 (interquartile range: IQR 5-10) and the median Acute Physiology and Chronic Health Evaluation (APACHE) II was 21 (IQR 16-25). The all-cause mortality score was 11.6%. The percentage of nutritional treatment initiated 24 h after ICU admission was 58%, with an EN of 34.2% and a parenteral nutrition (PN) of 16.0% in elderly patients who are critically ill. Patients who had gastrointestinal dysfunction with AGI stage from 2 to 4 were 25.2%. Compared to the survivors' group, the non-survivors group had a lower ratio of EN delivery (57% vs. 71%; p = 0.015), a higher ratio of post-pyloric feeding (9% vs. 2%; p = 0.027), and higher frequency of feeding interrupt (24% vs. 17%, p = 0.048). Multivariable logistics regression analysis showed that patients above 76 years old with OR (odds ratio) 2.576 (95% CI, 1.127-5.889), respiratory rate > 22 beats/min, and ICU admission for 24 h were independent risk predictors of the 28-day mortality study in elderly patients who are critically ill. Similarly, other independent risk predictors of the 28-day mortality study were those with an OR of 2.385 (95%CI, 1.101-5.168), lactate >1.5 mmol/L, and ICU admission for 24 h, those with an OR of 7.004 (95%CI, 2.395-20.717) and early PN delivery within 24 h of ICU admission, and finally those with an OR of 5.401 (95%CI, 1.175-24.821) with EN delivery as reference. This multi-center prospective study describes clinical characteristics, the mode and timing of nutrition treatment, frequency of AGI, and adverse effects of nutrition in elderly ICU patients. According to this survey, ICU patients with early PN delivery, older age, faster respiratory rate, and higher lactate level may experience poor prognosis.