OBJECTIVE: Our goal was to determine the false-negative and false-positive rates of antepartum testing by use of the modified biophysical profile. STUDY DESIGN: From Jan. 1, 1990, through Dec. 31, 1994, antepartum testing results were gathered prospectively and tabulated monthly. For 1 year, 1991, detailed intrapartum and neonatal data were collected from all women admitted and delivered as a result of an abnormal antepartum test result. RESULTS: The false-negative rate of the antepartum testing protocol was 0.8 per 1000 women tested. Sixty percent of those delivered because of an abnormal antepartum test had no evidence of short-term or long-term fetal compromise. False-positive test results led to preterm delivery in 1.5% of those tested before term. CONCLUSION: The false-negative rate of the modified biophysical profile is lower than that of the nonstress test and compares favorably with the false-negative rates of the contraction stress test and the complete biophysical profile. Iatrogenic prematurity resulting from intervention for false-positive test results occurred in 1.5% of women tested before 37 weeks. (AM J OBSTET GYNECOL 1996;174:812-7.)