OBJECTIVES: The primary aim of this study was to develop a noninvasive method for recording uterine contractile activity during pregnancy by systematically evaluating whether abdominal surface electromyography is representative of uterine electrical and contractile events. A secondary purpose was to use these techniques to determine when the uterus is in a state of preparedness for labor. STUDY DESIGN: Electrical activity was acquired by use of unipolar electrodes attached simultaneously to the uterine wall and to the abdominal surface of pregnant rats. Intrauterine pressure was recorded with a pressure transducer inserted in the uterine cavity. Computer-acquired records of electrical events (electromyography) and pressure were compared on different days of gestation, during spontaneous labor at term, or during preterm birth induced with an antiprogestin. Similarly, electrical activity was assessed after administration of agents that either stimulate (oxytocin) or inhibit (isoproterenol) contractility. Correlation analyses were performed between uterine electromyographic activity, surface electromyographic activity, and concomitant intrauterine pressure. The effects of vaginal wall stimulation were also evaluated at different times of pregnancy. RESULTS: The electrical activity recorded early in pregnancy (day 18) from the uterus consisted of irregular electrical bursts with little correspondence to the signals recorded from the surface (R² = 0.006). Later in gestation (days 19 through 21) the electrical activity of the uterus became more regular, consisting of frequent bursts with accordance between the signals recovered from the uterus and those collected from the surface (R² = 0.95). During labor (preterm or term) bursts recorded from either the uterus or the abdominal surface were of large amplitude and corresponded to ample changes in intrauterine pressure. Correlation coefficients between uterus and surface and uterus and pressure were R² = 1 and 0.96, respectively. Similarly, during preterm labor the coefficients were R² = 1 for uterine electromyographic activity versus surface electromyographic activity and R² = 0.99 for uterine electromyographic activity versus intrauterine pressure. Vaginal stimulation in early gestation was not followed by subsequent signal conduction to the uterus, whereas during delivery activity induced in the vagina propagated to the uterus and consequently to the abdominal surface. CONCLUSIONS: Abdominal surface recording of uterine electrical events are representative of the activity generated by the muscle cells of the uterus. During term and preterm labor uterine electrical activity and intrauterine pressure achieve maximum activity. Electromyographic monitoring with vaginal stimulation of the uterus may allow prediction of when the uterus is in a state required for labor. (AM J OBSTET GYNECOL 1996;174:744-53.)