PURPOSE--To test the hypothesis that the redundancy of the septum primum is more pronounced in fetuses with atrial premature beats than in normal fetuses. METHODS--Twenty-five consecutive fetuses with atrial premature beats detected by prenatal echocardiography as the sole alteration at a mean gestational age of 34 +/- 3.3 (26 to 38) weeks and a mean maternal age of 27 +/- 6.5 (18 to 39) years made up the study group. The control group was formed by 50 consecutive fetuses without rhythm disturbances and normal prenatal echocardiography, examined at a mean gestational age of 33.9 +/- 3 (26 to 39) weeks and a mean maternal age of 31 +/- 5.8 (18 to 43) years. A redundancy index (RI) was obtained as a ratio between the maximal septum primum excursion and the maximal left atrium diameter during atrial diastole, at the 4-chamber projection. Statistical analysis utilized the unpaired Student-Fisher's "t" test. RESULTS--The RI in the study group varied from 0.50 to 1.00 (mean = 0.758 +/- 0.127) and in the control group in ranged from 0.29 to 0.67 (mean = 0.449 +/- 0.080). These differences were statistically significant for p < 0.0001. CONCLUSION--Septum primum is more redundant in fetuses with atrial premature beats than in those with normal regular rhythm. Although it is still speculative that this feature may have a cause-effect relationship, the measurement of the RI during routine fetal echocardiography may allow detection of a potential risk factor FOR developing atrial arrhythmias.