Preterm delivery is associated with neonatal mortality, severe neonatal morbidity with long-term consequences, and disturbances in neurologic development. However, when prevention of preterm delivery is impossible, preventative administration of corticosteroids at least 24 hours prior to preterm delivery has been proved to reduce neonatal mortality and morbidity. Our aim was to find out why only 38.6% of children born between 28.0/7 and 32.6/7 weeks of gestation at the Ljubljana Maternity Hospital during the year 2000 received antenatal corticosteroids. An attempt was made to evaluate the standard values (sensitivity, specificity, and positive and negative predictive values) of the test that detects the phosphorylated form of insulin growth factor binding protein type 1 (phIGFBP-1) in the cervical fluid for prediction of preterm delivery, and to increase its usage. Our aim was also to draw up recommendations on how to increase the number of pregnant women with a high risk of preterm delivery that are given corticosteroids. It was expected that over 38.6% of pregnant women who delivered their babies between 28.0/7 and 32.6/7 weeks of gestation would have been treated with corticosteroids. However, this was not indicated in the delivery protocol. The desire was to evaluate standard values of the test for detecting phIGFBP-1. It is expected that our recommendations will increase the number of pregnant women at high risk of preterm delivery to be treated with corticosteroids. In the retrospective part, perinatal data were analyzed on women who delivered between 22.0/7 and 33.6/7 week of gestation in the year 2000 at the Ljubljana Maternity Hospital. It was analyzed how many were given corticosteroids, whether they received them at least 24 hours prior to delivery and whether this was marked in delivery protocol. In the prospective part of the research, phIGFBP-1 was determined in the cervical fluid of 48 women pregnant between 22.0/7 and 34.0/7 week of gestation who were either examined or hospitalized at the Department of Perinatology, Division of Obstetrics and Gynecology of the Ljubljana University Medical Centre during the period between 12 March 2002 and 13 February 2003. Two weeks after the test, its standard values were evaluated. The percentage of pregnant women who were given corticosteroids at the Ljubljana Maternity Hospital in the year 2000 was twice as high as data from the NPIS. 80% of women who delivered preterm between 22.0/7 and 33.6/7, and 84.5% of those who delivered between 28.0/7 and 32.0/7 week of gestation were treated with corticosteroids. Along with other risk factors, determination of phIGFBP-1 is useful in the prediction of preterm delivery. According to our research performed on a low number of patients, its sensitivity is 100%, specificity 59%, positive predictive values 15% and negative predictive values 100%. It is suggested that an objective determination of the risk factors for preterm delivery should be performed at the first examination of a pregnant women and later, throughout the entire pregnancy instead of the diagnosis of threatened preterm delivery, using phIGFBP-1 for high negative predictive values and administering corticosteroids when there is a high risk of preterm delivery.