To investigate whether children with concurrent wasting and stunting require therapeutic feeding and to better understand whether multiple diagnostic criteria are needed to identify children with a high risk of death and in need of treatment. Community-based cohort study, following 5751 children through time. Each child was visited up to four times at 6-month intervals. Anthropometric measurements were taken at each visit. Survival was monitored using a demographic surveillance system operating in the study villages. Niakhar, a rural area of the Fatick region of central Senegal.ParticipantsChildren aged 6-59 months living in thirty villages in the study area. Weight-for-age Z-score (WAZ) and mid-upper arm circumference (MUAC) were independently associated with near-term mortality. The lowest WAZ threshold that, in combination with MUAC, detected all deaths associated with severe wasting or concurrent wasting and stunting was WAZ <-2·8. Performance for detecting deaths was best when only WAZ and MUAC were used. Additional criteria did not improve performance. Risk ratios for near-term death in children identified using WAZ and MUAC suggest that children identified by WAZ <-2·8 but with MUAC≥115 mm may require lower-intensity treatment than children identified using MUAC <115 mm. A combination of MUAC and WAZ detected all near-term deaths associated with severe anthropometric deficits including concurrent wasting and stunting. Therapeutic feeding programmes may achieve higher impact if WAZ and MUAC admission criteria are used.