Background: There is an increasing recognition of malnutrition among infants under 6 mo of age (U6M). Current diagnosis criteria use weight-for-length z scores (WLZs), but the 2006 WHO standards exclude infants shorter than 45 cm. In older children, midupper arm circumference (MUAC) predicts mortality better than does WLZ. Outcomes may also be influenced by exposure to HIV and size or gestational age at birth. Diagnostic thresholds for WLZ, MUAC, and other indexes have not been fully evaluated against mortality risk among U6M infants.Objective: The aim was to determine the association of anthropometric indexes with risks of inpatient and postdischarge mortality among U6M infants recruited at the time of hospitalization.Design: We analyzed data from a cohort of U6M infants admitted to Kilifi County Hospital (2007-2013), Kenya. The primary outcomes were inpatient death and death during follow-up over 1 y after discharge. We calculated adjusted RRs for inpatient mortality and HRs for postdischarge mortality for different anthropometric measures and thresholds. Discriminatory value was assessed by using receiver operating characteristic curves.Results: A total of 2882 infants were admitted: 140 (4.9%) died in the hospital and 1405 infants were followed up after discharge. Of these, 75 (5.3%) died within 1 y during 1318 child-years of observation. MUAC and weight-for-age z score (WAZ) predicted inpatient and postdischarge mortality better than did WLZ (P < 0.0001). A single MUAC threshold of <11.0 cm performed similarly to MUAC thresholds that varied with age (all P > 0.05) and performed better than WLZ <-3 for both inpatient and postdischarge mortality (both P < 0.001). Reported small size at birth did not reduce the risk of death associated with anthropometric indexes.Conclusions: U6M infants at the highest risk of death are best targeted by using MUAC or WAZ. Further research into the effectiveness of potential interventions is required.