This paper examines the relationship between the standard deviation scores of weight-for-age (WAZ), height-for-age (HAZ) and weight-for-height (WHZ) calculated using the CDC/WHO reference. The pattern of WHZ with age is explored for CDC/WHO reference children following set paths for WAZ and HAZ, and the nature of the regression of WAZ on HAZ and WHZ is examined for the Ghana Demographic and Health Survey data (1988). Reference children with heights and weights at the lower end of the range for their population showed discrepancies in their WHZ indices. A marked discontinuity between 23 and 24 months reflected the junction between the two data sets from which the CDC/WHO reference was derived. There was also clear curvature in the values for WHZ between 3 and 23 months, particularly apparent for children with weights and heights well below the reference median. While the regression equation for the Ghanaian data was similar to that already described by Keller in 1983, marked curvature was observed in the residual plot for children under 2 years. Consistency with age is a prerequisite for a weight-for-height index. The peculiarities of this index as based on the CDC/WHO reference undermine the validity of its use in measuring malnutrition among young children in developing countries. It is suggested that serious consideration is given to the use of an alternative index of wasting and to the development of alternative reference figures.