Background: Stress and anxiety are common in pregnancy and shown to have adverse effects on maternal and infant health outcomes. The aim of this review and meta-analysis was to assess the effectiveness of music-based interventions in reducing levels of stress or anxiety among pregnant women. Methods: Six databases were searched using key terms relating to pregnancy, psychological stress, anxiety and music. Inclusion criteria were randomised controlled or quasi-experimental trials that assessed the effect of music during pregnancy and measured levels of psychological stress or anxiety as a primary or secondary outcome. Two authors independently assessed and extracted data. Quality assessment was performed using The Cochrane Collaboration risk of bias criteria. Meta-analyses were conducted to assess stress and anxiety reduction following a music-based intervention compared to a control group that received routine antenatal care. Results: Five studies with 1261 women were included. Music interventions significantly reduced levels of maternal anxiety (Standardised Mean Difference (SMD): -0.21; 95% Confidence Interval (CI) -0.39, −0.03; p = 0.02). There was no significant effect on general stress (SMD: -0.08; 95% CI -0.25, 0.09; p = 0.35) or pregnancy-specific stress (SMD: -0. 02; 95% CI -0.19, 0.15; p = 0.80). The methodological quality of included studies was moderate to weak, all studies having a high or unclear risk of bias in allocation concealment, blinding and selective outcome reporting. Conclusions: There is evidence that music-based interventions may reduce anxiety in pregnancy; however, the methodological quality of the studies was moderate to weak. Additional research is warranted focusing on rigour of assessment, intensity of interventions delivered and methodological limitations.