Evidence on adverse effects of maternal macrolide use during pregnancy is inconsistent. We conducted a systematic review and meta-analysis to investigate the association between macrolide use during pregnancy and adverse fetal and child outcomes. We included observational studies and randomised controlled trials (RCTs) that recorded macrolide use during pregnancy and child outcomes. We prioritized comparisons of macrolides with alternative antibiotics (mainly penicillins or cephalosporins) for comparability of indication and effect. Random effects meta-analysis was used to derive pooled odds ratios (OR) for each outcome. Subgroup analyses were performed according to specific types (generic forms) of macrolide. Of 11,186 citations identified, 19 (10 observational, 9 RCTs) studies were included (21 articles including 228,556 participants). Macrolide prescribing during pregnancy was associated with an increased risk of miscarriage (pooled ORobs 1·82, 95% CI 1·57-2·11, three studies, I2 = 0%), cerebral palsy and/or epilepsy (ORobs 1·78, 1·18-2·69; one study), epilepsy alone (ORobs 2·02, 1·30-3·14, one study; ORRCT 1.03, 0.79-1.35, two studies), and gastrointestinal malformations (ORobs 1·56, 1·05-2·32, two studies) compared with alternative antibiotics. We found no evidence of an adverse effect on 12 other malformations, stillbirth, or neonatal death. Results were robust to excluding studies with high risk of bias. Consistent evidence of an increased risk of miscarriage in observational studies and uncertain risks of cerebral palsy and epilepsy warrant cautious use of macrolide in pregnancy with warnings in drug safety leaflets and use of alternative antibiotics where appropriate. As macrolides are the third most commonly used class of antibiotics, it is important to confirm these results with high quality studies.