Sao Paulo Med J 2003; 121(4):147-148. The use of anthropometric references for the evaluation of intrauterine growth has shown that children whose growth was restricted are more predisposed to metabolic disturbances during the neonatal period, alterations in somatic and neurocognitive development during infancy, increased morbidity and mortality in the first years of life and the appearance of chronic non-transmissible diseases during adulthood. Birth weight has been the index most used for assessment of intrauterine growth. Weight is not an ideal measurement for evaluating growth: it is merely easier to measure, and the balances now available can do this to a precision of five grams. The weight gives an assessment of all the tissues together, and greater weight does not necessarily signify good growth: it may be achieved at the cost of liquid retention or fat deposition. On the other hand, the measurement of length requires a refined technique, with an appropriate instrument, and it is not always easy among newborns. Other measurements and the relationships between these have been utilized in such evaluations: cranial perimeter, skinfolds, mid-arm circumference in association with cephalic perimeter and skinfolds, and mid-arm circumference and muscle area. All these measurements have technical or interpretative limitations on their results, which makes it difficult to use them routinely in clinical practice. The evaluation of weight distribution at birth is more adequate when it is associated with gestational age, as has been widely done since Lubchenco (1963) published his intrauterine growth curves. Methodological differences and the prevalence of factors known to be associated with reduced birth weight may explain the large differences seen between this and other curves developed subsequently. The observation that not all babies born with low weight are born before the full gestational term led, in the words of Wilcox (2001), to the invention of a new disease from the 1970s onwards: intrauterine growth restriction (IUGR). The usual definition of IUGR is “small for the gestational age”, i.e. the smallest 10% of each stratum of gestational age. The determination of gestational duration is also a problem in daily clinical practice. It is usually obtained via information from the mother regarding the date of the last menstruation, which brings the possibility of various orders of error, from the women who cannot remember the date to those who present Neonatal anthropometry and neonatal outcome