Undernourishment has been associated with poor outcomes of critical illness in children. The effects of withholding parenteral nutrition (PN) for 1 week in undernourished critically ill children are unknown. To assess the outcome effects of withholding PN for 1 week in undernourished critically ill children. This is a subanalysis of the randomized clinical trial Pediatric Early vs Late Parenteral Nutrition in Intensive Care Unit (PEPaNIC; N = 1440), which focused on the subgroup of pediatric intensive care unit (PICU) patients identified as undernourished on admission. Children included in the PEPaNIC trial were enrolled between June 18, 2012, and July 27, 2015. Undernourishment was defined as weight-for-age z score less than -2 in children younger than 1 year, and body mass index-for-age z score less than -2 in children 1 year or older. Data analysis was conducted from August 3, 2017, to July 6, 2018. Patients were randomized to initiation of supplemental PN within 24 hours (early PN) or after 1 week (late PN) when enteral nutrition was insufficient. Primary end points were risk of new infections acquired in the PICU and time to live PICU discharge, assessed via multivariable logistic regression and Cox proportional hazard analyses, adjusted for risk factors. A total of 289 of 1440 children (20.1%), term newborn to age 17 years, were identified as undernourished, of whom 150 of 717 patients (20.9%) were in the late PN group and 139 of 723 patients (19.2%) were in the early PN group. On admission, characteristics were similar among the treatment groups. Mean (SD) weight z scores were -3.33 (1.18) in the late PN group and -3.21 (1.09) in the early PN group. Compared with well-nourished PICU patients, undernourishment on admission was associated with lower likelihood of an earlier live PICU discharge (adjusted hazard ratio, 0.86; 95% CI, 0.75-0.99; P = .03). Among undernourished PICU patients, late PN reduced the risk of new infections by 11.0% (adjusted odds ratio, 0.39; 95% CI, 0.19-0.78; P = .01), and shortened the duration of PICU stay by a median of 2 days (earlier live PICU discharge: adjusted hazard ratio, 1.37; 95% CI, 1.06-1.75; P = .01). The safety outcomes mortality, incidence of hypoglycemia during the first week, and incidence of weight deterioration during PICU stay were similar between the treatment groups. In undernourished critically ill children, withholding PN for 1 week was clinically superior to early PN. ClinicalTrials.gov Identifier: NCT01536275.