This study sought to develop a simple echocardiographic predictor of persistent left (L) superior vena cava (SVC) in subjects with bilateral SVC. Two groups of children were studied: one with known LSVC (n = 19) and the other, a control group, without LSVC (n = 15). Both groups were of similar age (3.5 +/- 3.0, mean +/- SD vs 3.8 +/- 3.1 years; p = 0.8) and weight (14.6 +/- 7.1 vs 15 +/- 8 kg; p = 0.9). The left innominate vein was either absent (n = 11) or hypoplastic (n = 8) in the LSVC group. The ratio of the innominate vein to the innominate artery was found to be independent of age or body surface area but was significantly smaller in the LSVC group than in the control group (0.33 +/- 0.1 vs 0.97 +/- 0.1; p < 0.001). A cutoff value of 0.47 or less discriminated the LSVC group from the control subjects. Interobserver and intraobserver variations, although important, did not influence the discriminating value of the ratio in the diagnosis of LSVC. Validation of the proposed ratio in 30 consecutive prospectively studied patients with LSVC proved to be 100% sensitive in predicting LSVC. No false-positive diagnosis of LSVC was made when this principle was applied. Absent or hypoplastic left innominate vein measuring 0.47 or less of the innominate artery is an easily recognizable and reliable echocardiographic predictor of LSVC.