Objective Evaluate the strength of association and predictive value of measurements of amniotic fluid volume (AFV) for small for gestational age and compromise of fetal wellbeing.

Design Systematic literature review with random effects meta-analysis to compute summary odds ratios (OR) to assess prognostic association and assess predictive ability with sensitivity, specificity and likelihood ratios. Study characteristics, design, methodological and reporting quality were objectively assessed.

Data Sources and Eligibility Systematic search (inception to October 2011) for studies comparing AFV measures and outcomes of fetal size or wellbeing.

Results 43 studies reporting 244,493 fetuses included. Strong associations between oligohydramnios (heterogenic definition) and birth weight < 10th centile [OR 6.31 (4.15–9.58)] in a high risk population (6 studies, 28510 fetuses), and mortality [Neonatal death OR 8.72 (2.43–31.26) 6 studies, 55735 fetuses, and perinatal mortality in a high risk population OR 11.54 (4.05–32.9) 2 studies, 27891 fetuses] were identified. There was no significant association between oligohydramnios and abnormal cord pH or adverse perinatal outcome.

There was no significant association between polyhydramnios (heterogenic definitions) and poor fetal growth [Birth weight < 10^th centile OR 0.37 (0.07–1.95)]. A strong association between polyhydramnios (maximum pool depth > 8 cm or amniotic fluid index ≥ 25 cm) and birth weight > 90^th centile [OR 11.41 (7.09–18.36) 1 study, 3960 fetuses] was found. Despite strong associations predictive ability was poor with significant heterogeneity despite sub-group analysis.

Conclusion Oligohydramnios is associated with poor fetal growth and mortality. Polyhydramnios is associated with BW > 90th centile. Despite strong associations, overall predictive ability was poor. MPD accuracy was slightly improved over AFI and is the recommended technique.