OBJECTIVE To compare the efficacy in detecting signs of fetal hypoxia in labour of intermittent (I-group) versus continuous (C-group) electronic fetal monitoring in women with low or moderate risk factors for fetal distress. METHODS A prospective, randomised study. METHODS A tertiary referral centre. METHODS Four thousand and forty-four parturients at low risk for obstetric complications with a reactive fetal heart rate admission test at arrival in labour. During the study period (October 5 1989 to May 31 1991), 5647 women were delivered in the labour ward. Of these, 1178 women (20.9%) were excluded because of high risk factors in pregnancy or at admission for labour, including women undergoing elective caesarean section. Of the remaining 4469 women 4044 (90.5%) were randomised to either intermittent (n = 2015) or continuous monitoring (n = 2029) during the first stage of labour. METHODS In the C-group the fetal heart rate was recorded continuously with electronic fetal monitoring during the first stage of labour. In the I-group the fetal heart rate was recorded with electronic fetal monitoring for 10 to 30 min every 2 to 2.5 h during the first stage of labour, and the fetal heart rate was auscultated every 15 to 30 min in between recording periods. If complications occurred, recording was changed to continuous. In the second stage of labour all the women were monitored continuously. Umbilical cord artery acid-base status was assessed at birth. METHODS Duration of electronic fetal monitoring, rates of abnormal fetal heart rate patterns, caesarean section for fetal distress, acidosis in umbilical cord arterial blood at birth, Apgar scores of less than 7 at 1 or 5 min, and referrals to the neonatal intensive care unit. RESULTS There were no significant differences between the study groups in the incidence of ominous fetal heart rate recordings: 6.3% (I-group) versus 6.6% (C-group), or the interval from arrival to first detected abnormal fetal heart rate, although the number of suspicious fetal heart rate patterns was higher in the C-group (28.6%) than in the I-group (24.6%). In the I-group electronic fetal monitoring was performed for (median monitoring time) 38.8% of the first stage of labour as compared with 78.6% in the C-group. The incidence of caesarean section for fetal distress was similarly low in both groups: 1.2% versus 1.0%. There were no significant differences in the immediate neonatal outcome in terms of umbilical artery pH, Apgar scores, or admissions to the neonatal care unit. CONCLUSIONS Intermittent use of electronic fetal monitoring at regular intervals (with stethoscopic auscultation in between) appears to be as safe as continuous electronic fetal monitoring in low risk labours.