CONTEXT: Intrauterine growth curves are extremely useful for classifying newborn children and predicting neonatal diseases. However, such curves rely on knowledge of the gestational age, which is not always easily obtained. Therefore, the study of other anthropometric measurements and their interrelationship is always desirable, in order to attain such objectives. OBJECTIVE: To evaluate whether newborns' birth weight and length can identify neonatal diseases, independent of knowledge of the gestational age. TYPE OF STUDY: Retrospective study. SETTING: Institute of Teaching and Research of Hospital Israelita Albert Einstein, São Paulo, Brazil. PARTICIPANTS: During the period from February 1995 to January 1998, 8,397 live newborns were studied in the hospital's maternity ward. PROCEDURES: The weight and length of live newborns were obtained at birth, thus allowing the analysis of weight-for-length adequacy, i.e. the distribution of birth weight for each class of birth length. These measurements were determined for the first 4,634 live newborns and the 10th and 90th percentiles were established. These parameters were applied to the next 3,763 consecutive newborns of the same population. The relationships between these variables and some neonatal diseases were investigated. The significance level adopted was p<0.05. MAIN MEASUREMENTS: Birth weight and length, weight-for-length adequacy (10th and 90th percentiles for weight distribution in each 1-cm length class), weight/length index (10th and 90th percentiles of newborn's weight divided by the length) and frequent neonatal diseases in this population. RESULTS: There was a significant association of adequacy and index with the following affections: asphyxia, jaundice, hypoglycemia, hypomagnesemia, congenital pneumonia, pulmonary hypertension and sepsis. Additionally, there was a relationship between the index and respiratory distress syndrome, transient tachypnea and persistent ductus arteriosus. CONCLUSIONS: Weight-for-length adequacy and weight/length index alone, without the knowledge of gestational age, were able to identify newborns at risk for some selected neonatal diseases.