Abstract: Determination of Pb isotopes in melt inclusions by Laser Ablation -Multiple Collector-Inductively Coupled Plasma-Mass Spectrometry (LA-MC-ICP-MS) is limited to samples formed later than Mesozoic because U and Th signals are not monitored. In this study, Pb isotopic compositions of two reference glasses and ancient melt inclusions which have undergone significant U-Th decay are determined by LA-MC-ICP-MS. The effect of the U/Pb and Th/Pb analytical errors on the calculation of initial Pb isotopic ratios was also evaluated. In the experiment USGS NKT-1G is used as the external calibrant and a standard-sample-standard bracketing method is used to correct instrument drift and mass bias. Analytical precisions of ²⁰⁸Pb/²⁰⁶Pb and ²⁰⁷Pb/²⁰⁶Pb of USGS standards, BHVO-2G and TB-1G, are better than 0.30% (2RSD) and both agree within 0.30% of the preferred values. However, ²³²Th/²⁰⁶Pb and ²³⁸U/²⁰⁶Pb ratios are highly variable and yield an external precision of 5%. Error propagation calculations indicate that the ages of the samples have significant effects on the errors of the initial lead isotopic ratios. For Phanerozoic samples that are earlier than 540 Ma, even the differences between the measured ²³²Th/²⁰⁶Pb and ²³⁸U/²⁰⁶Pb ratios and the real values reach 10%, yet the age-corrected initial lead isotopic ratios are still within 0.80% deviations relative to the real values. Therefore, the proposed method in this study can determine the lead isotopic compositions of melt inclusions in samples from Cenozoic to Paleozoic, expanding the geological application.