background The lead isotope ratios (LIR) differ among different sourced samples. Previous domestic and oversea studies on source tracing by LIR in human blood or urine mainly focused on the comparison of blood or urine samples from the same or different individuals, while few comparisons between biological and environmental samples, and the reported relative standard deviations (RSDs) of the main LIR (^207/206Pb and ^208/206Pb) fluctuate widely from 0.3% to 1%. Objective To optimize inductively coupled plasma mass spectrometry (ICP-MS), obtain a better RSD, and determine LIRs of human blood, urine, and related environmental samples. Methods The ICP-MS was optimized for operating conditions and parameters according to the sensitivity and RSD of LIR. The study subjects were 40 lead-exposed workers in a lead-acid battery factory and 2 lead poisoned children in a hospital. The samples included 40 blood and 40 urine samples from the workers before shift, 4 dust samples and 2 water samples in the workplace on the same day before shift, 2 blood and 3 urine samples from the children before hospital admission due to lead-poisoning, and 4 urine samples after medical treatment. After heating and acid digestion, the LIR (^207/206Pb and ^208/206Pb) of biological and environmental samples were determined by the optimized ICP-MS method. t-test and two-dimensional traceability graphics were adopted to analyze the detection results. Results The calibrated RSDs of the LIR (^207/206Pb and ^208/206Pb) of lead isotope standard solution were 0.11% and 0.08% respectively, and the NIST-SRM-981 actual values were 0.91531±0.00097 and 2.1670±0.0017, respectively. When the total concentration of lead was greater than 5 μg·L⁻¹, the RSD of each isotope ratio was stable gradually; when the total concentration of lead was between 10-80 μg·L⁻¹, the RSD was below 0.20%. There were statistically significant differences in the blood and urine LIR (^207/206Pb and ^208/206Pb) of the lead-exposed workers (t=5.831, P<0.001; t=21.021, P<0.001), the LIR (^207/206Pb and ^208/206Pb) between workplace dust samples and workers’ urine samples (t=−6.879, P=0.038; t=12.521, P<0.001), and the ^208/206Pb between workplace dust samples and workers’ blood samples (t=−10.46, P<0.001), except the ^207/206Pb between workplace dust samples and workers’ blood samples (t=−0.12, P=0.912). In the patients afflicted with lead poisoning, the projection points of LIR of blood and urine samples from the same individual were not at the same level in the two-dimensional model, nor was the LIR of urine samples before and after medical treatment of the same individual. Conclusion The optimized ICP-MS can control the RSD of main LIR (^207/206Pb and ^208/206Pb) below 0.20%. There are differences in the LIR distributions of different samples.