The preterm and postterm delivery rates and the percentage of infants with intrauterine growth retardation are dependent on the gestational age recorded at delivery. At our institution a sharp increase in the preterm delivery rate and a coincident decrease in the postterm delivery rate and the rate of intrauterine growth retardation were noted. Over a 3-year period, while the characteristics of the obstetric population changed only slightly, the gestational age distribution shifted, with a decrease in the mean gestational age of about 1 week and a risk in the reported preterm delivery rate from 12% to 17%. About 15% of this rise was explained by an increase in obstetric interventions, and another 15% can be explained by changes in the way physicians rounded off gestational age. The majority of the increase in the preterm delivery rate was related to changes involving ultrasonographic examinations. These changes included a greater percentage of the population examined, trends toward earlier examinations, a tendency for the physicians to use ultrasonography rather than the last menstrual period in choosing the final gestational age, the use of different standards, an increase in the number of structures measured, and the weight given to various structures for determination of gestational age. It is apparent that changes in use of ultrasonography had a profound effect on the reported gestational age distribution at our institution.