BACKGROUND: Determining the significance of a patent ductus arteriosus (PDA) remains difficult in preterm neonates and current echocardiographic markers give little information about ductal steal. We hypothesized that the relationship between celiac artery flow (CAF) and left ventricular output (LVO) in the presence of a PDA may determine haemodynamic significance. OBJECTIVES: To examine CAF to LVO ratio (CAF:LVO) in the presence and the absence of a PDA in preterm neonates, and to compare CAF:LVO to current echocardiographic markers of a significant PDA. METHODS: This was a prospective observational study of neonates <1,500 g. Echocardiography was performed at 12 h and day 3. PDA, LVO and CAF were measured by echocardiography. The infants were divided into those who developed a significant PDA (PDA >1.4 mm and left atrial to aortic ratio >1.5 on day 3) and controls. A further assessment was carried out following successful PDA treatment with ibuprofen or surgical ligation. RESULTS: A total of 33 infants were enrolled. 19 infants had a PDA (median gestation 27 weeks, birth weight 915 g), and 14 controls (gestation 28.7 weeks, birth weight 1,110 g). At 12 h, there was no difference in CAF, LVO or CAF:LVO. On day 3, there was a significant difference in CAF:LVO between PDA versus control groups. CAF:LVO significantly correlated with conventional markers of ductal significance. A CAF:LVO value of 0.10 had a sensitivity of 93% and a specificity of 78% for the presence of a significant PDA. Following successful PDA treatment CAF:LVO returned to levels similar to controls. CONCLUSION: CAF:LVO may be used to determine which PDA warrants treatment and serve as a marker of treatment success. Trials are needed to correlate CAF:LVO with outcome and the impact of basing PDA treatment on this ratio.