OBJECTIVES: The objectives of this study were to estimate the accuracy of using mid-upper-arm circumference (MUAC) measurements to diagnose severe wasting by comparing the new standards from the World Health Organization (WHO) with those from the US National Center for Health Statistics (NCHS) and to analyze the age independence of the MUAC cutoff values for both curves. METHODS: We used cross-sectional anthropometric data for 34,937 children between the ages of 6 and 59 months, from 39 nutritional surveys conducted by Doctors Without Borders. Receiver operating characteristic curves were used to examine the accuracy of MUAC diagnoses. MUAC age independence was analyzed with logistic regression models. RESULTS: With the new WHO curve, the performance of MUAC measurements, in terms of sensitivity and specificity, deteriorated. With different cutoff values, however, the WHO standards significantly improved the predictive value of MUAC measurements over the NCHS standards. The sensitivity and specificity of MUAC measurements were the most age independent when the WHO curve, rather than the NCHS curve, was used. CONCLUSIONS: This study confirms the need to change the MUAC cutoff value from <110 mm to <115 mm. This increase of 5 mm produces a large change in sensitivity (from 16% to 25%) with little loss in specificity, improves the probability of diagnosing severe wasting, and reduces false-negative results by 12%. This change is needed to maintain the same diagnostic accuracy as the old curve and to identify the children at greatest risk of death resulting from severe wasting.