BACKGROUND/OBJECTIVES: The precise role of fish oil-based lipid emulsions (FBL=any lipid emulsion containing any amount of fish oil) and soy oil-based lipid emulsions (SBL=any lipid emulsion not containing fish oil or containing at least 50% soy oil) in neonatal parenteral nutrition (PN) needs to be explored. SUBJECTS/METHODS: The PubMed, MedLine and Google Scholar databases were searched for randomised control trials in which PN with FBL and SBL was the only difference between intervention and control groups. Methodological quality was assessed based on the Cochrane handbook for systemic reviews and Jadad's score. Revman 5.2 software was used for meta-analysis. RESULTS: Of 420 trials, 25 met the inclusion criteria. The aggregate results showed that both FBLs and SBLs are beneficial. FBLs were associated with significantly lower incidences of cholestasis (relative risk (RR)=0.50; 95% confidence interval (CI)=0.27-0.92, P=0.03) compared with SBLs. FBLs also had a tendency to improve the rates of weight gain, increase in head circumference and the time to regain birth weight (mean difference (MD)=0.24; 95% CI=-0.63-0.15). SBLs were associated with reduced duration of respiratory support (MD=2.22, 95% CI=0.35-4.09, P=0.02) and mortality rate in the first 28 days (RR=1.24, 95% CI=-0.72-2.13). CONCLUSIONS: Both FBLs and SBLs are beneficial for neonatal PN. FBLs reduce cholestasis, and SBLs reduce the duration of respiratory support.