Uterine activity reported by the subjects was con- firmed by external uterine tocodynamometry. Evidence of vaginal bleeding was visually assessed during the pelvic examination. Two vaginal specimens were ob- tained for fetal fibronectin assay, one from the external cervical OS and the second from the posterior fornix of the vagina, by means of a Dacron swab.’ The collected specimens were extracted in an antiprotease buffer containing bovine serum albumin and stored at 4” C until assay. Concentrations of fetal fibronectin were measured by an enzyme-linked immunosorbent assay that uses the murine monoclonal antibody FDC-6.3 This antibody specifically distinguishes fetal from hepatic fibronectin by recognizing a unique glycosylhexapep- tide structure within the variably inserted IIICS region of fibronectin.4, 5 The results were considered positive when both specimens had a fetal fibronectin concentra- tion > 50 rig/ml, as determined by Lockwood et aL3 Continuous variables were analyzed by nonparamet- ric tests. Categoric data were analyzed with either Fish- er’s exact test or the x2 test with continuity correction as indicated by cell frequencies. A stepwide multiple logis- tic regression analysis was also performed to assess the association of the independent variables (fetal fibronec- tin, cervical dilatation, and uterine activity) to the de- pendent variable (delivery < 7 days). The primary end points were delivery < 7 days, the frequency of delivery before 37 weeks, and the interval from sampling to delivery for fibronectin-positive and fibronectin-nega- tive subjects. We calculated that 188 subjects would be required to demonstrate a significant association be- tween fetal fibronectin and outcome with a positive predictive value of 50% and a negative predictive value of 80%. Comparisons of fetal fibronectin were also made to other diagnostic criteria for preterm labor including cervical dilatation > 1 cm, contraction fi-e- quency, and vaginal bleeding. Use of tocolytic medica- tion and mean birth weight were also compared for fetal fibronectin positive and fetal fibronectin-negative subjects.