Serial electrocardiograms (ECGs) were studied prospectively in 80 apparently healthy newborn infants; 30 infants exposed in utero to prolonged tocolytic therapy (21 to ritodrine and 9 to isoxsuprine) and 50 infants non-exposed in utero to drugs (control group) matched for gestational age, Apgar score, and birth weight. Duration of exposure to tocolysis was at least 30 days (30-180 days) with an oral dosage of 10 mg 3 times daily. ECGs were graded for changes suggestive of ischaemia using the arbitrary grading system described by Jedeikin et al. In all infants with ECG features of myocardial ischaemia, serum creatine-phosphokinase iso-enzyme (CK-MB) activity was measured. Six out of 21 infants to ritodrine and six out of nine infants exposed to isoxsuprine showed a degree of ECG ischaemia which persisted for several weeks. No control infant presented grade 2 or 3 ECG changes after the 5th day of life. The results of this study seem to show that prolonged tocolytic therapy with beta-sympathomimetics has side-effects on the fetal myocardium and suggest that this treatment be reserved only for selective cases and/or for short periods of time.