ests for genuine preterm labor would not allow subjects to be treated, but such a trial would be extremely difficult if not impossible in most centers. Our design allowed subjects to receive tocolysis on the basis of traditional criteria of a change in cervical dilatation and persistent uterine activity. It is interesting to note that the use of tocolytic agents did not differ significantly between fetal fibronectin-positive and fetal fibronectinnegative patients. A clinical decision to treat with tocolytics in this population is actually a direct indicator of the clinician’s best judgment about the presence or absence of preterm labor. That judgment presumably includes cervical dilatation, effacement, consistency, the number and intensity of uterine contractions, and the patient’s obstetric history. In this multicenter university study the presence or absence of fetal fibronectin was superior to a diagnosis of preterm labor on the basis of traditional criteria. Fetal fibronectin performed espe-