Objectives.To determine the reliability of respiratory rate and subcostal retractions in diagnosing acute lower respiratory infection (ALRI) in undernourished children.Methods.Three hundred twelve children with ALRI and 446 with upper respiratory infection were classified according to weight and height as normal, stunted, wasted or stunted and wasted and also as normal, underweight or marasmus. The sensitivity and specificity of tachypnea, subcostal retractions and the presence of either sign in identifying children with a clinical diagnosis of ALRI or radiologic pneumonia in each of the nutritional categories were determined and compared.Results.Among children with ALRI the mean respiratory rate in those with normal nutrition (61.5 ± 16.1,n= 160) was not significantly different from those who were stunted (57.5 ± 16.5,n= 59), wasted (61.3 ± 14,n= 66) or stunted and wasted (55.4 ± 12.8,n= 27) (P> 0.05) or from those classified as underweight (60 ± 15.9,n= 150) or marasmus (62.5 ± 14.5,n= 27) (P> 0.4). The sensitivity and specificity of tachypnea, subcostal retraction or the presence of either sign in detecting ALRI was also not statistically significantly different among the children in the different nutritional categories (P> 0.05). The sensitivity of tachypnea or subcostal retraction in identifying children with radiologic pneumonia was also not significantly different among children in the different nutritional categories; the sensitivity of either sign was higher in under-weight children than in children with normal nutrition (P= 0.028).Conclusions.The data suggest that the current WHO algorithm is suitable for diagnosis of ALRI in undernourished children.