ObjectiveKnowledge of alcohol consumption in pregnancy is important for early identification of children with fetal alcohol spectrum disorder. We investigated whether alcohol biomarkers fatty acid ethyl esters (FAEEs) and ethyl glucuronide (EtG) in meconium are predicted by maternal or newborn demographics and/or correlate with confidential early postnatal self-report of alcohol consumption in pregnancy.DesignAnonymised, observational population-based study.SettingInner-city maternity unit, Glasgow, UK.PatientsSingleton mother/infant dyads delivering every fourth day.InterventionsMother: confidential postnatal interview. Baby: meconium sample for FAEEs and EtG.Results840/908 mothers consented. 370 (46.4%) reported alcohol consumption in pregnancy, generally of modest amount; for 114 (13.6%) this was after 20 weeks’ gestation. Alcohol consumption in later pregnancy was more commonly reported by older (31.3 vs 29.5 years) women of white British ethnicity (p<0.05); their babies were on average 118 g heavier (p=0.032). FAEEs were identified in all meconium samples; concentration was ≥600 ng/g in 39.6%. EtG concentration was ≥30 ng/g in 14.5%. Neither biomarker was associated with maternal age, body mass index or socioeconomic status but when EtG was ≥30 ng/g, the mother was less likely to identify as white British (71.3% vs 81.8%, p=0.028). Sensitivities of FAEEs ≥600 ng/g and EtG ≥30 ng/g were 43.1% and 11.6%, respectively for postnatal self-report of alcohol use in later pregnancy (specificities 60.6% and 84.8%).ConclusionsFAEEs and EtG measured in meconium have low sensitivity and specificity for self-reported alcohol consumption after 20 weeks’ gestation in an unselected Scottish population.