Four issues in the use and interpretation of anthropometry are discussed at the level of the population and of the individual. The first issue is the index or indices of choice: weight-for-height versus height-for-age versus weight-for-age. The selection of an index or indices depends upon many factors, and no one index is completely adequate in all situations. Proposed criteria are provided to assess the severity of low anthropometry within populations. The second issue is the scale of the index: z-scores (or standard deviations) versus percentiles versus percent-of-median. z-Scores have several properties that make them superior to the other two scales. A third issue deals with limitations in the current growth reference; one of these is the disjunction between the growth curves at 2 years of age, resulting from the use of two different populations in the reference. It is important that this disjunction be recognized by researchers so that the anthropometric findings are interpreted correctly for this age range. Lastly, some issues to do with the collection of single versus multiple anthropometric measurements on children are discussed.