Introduction: Polyhydramnios is common; the majority of cases are idiopathic, but maybe associated with fetal abnormality. Literature suggests the volume of amniotic fluid discriminates idiopathic from pathological polyhydramnios but is not unanimous. We assessed fetal anomaly incidence amongst women with polyhydramnios and the role of discriminatory variables in identifying pathological cases. Methods: Retrospective observational cohort study at an inner-city London fetal medicine centre. Records for patients referred and/or diagnosed with polyhydramnios were reviewed as well as maternal/fetal demographics, amongst singleton pregnancies using the Astraia™ database from January 2015-2016. Estimated fetal weight was calculated using the Hadlock model (biometry undertaken at diagnosis). Student's t-test/one-way ANOVA compared means; chi-squared tests compared proportions. Results: 120 cases were identified. 36 (30%) had fetal abnormality. There was no difference in AFI between fetuses with an abnormality and without (26.7 vs 25.2 cm, P = 0.22). AFI was normalised for weight (AFI (cm)/estimated fetal weight (kg)): AFI/kg was significantly different between cases with fetal abnormality and without (24.4 vs 16.7 cm/kg, P < 0.001) - incidence of abnormality increased with increasing AFI/kg (P = 0.007). Early gestational diagnosis was associated with higher rates of anomaly (P = 0.004). Differences in AFI/kg between those with and without abnormality were not significant when adjusted for gestation. AFI was significantly higher in cases of abnormality diagnosed at later gestation (P = 0.005). Conclusion: Excess volume of amniotic fluid alone does not denote abnormality. Earlier gestations and higher AFI/kg corresponded with significantly increased rates of anomaly. However, the latter is a result of confounding by gestation, which is closely correlated with fetal weight.