The standard addition (SA) method has been successfully employed in measuring metal-stable isotopes in recent years. Here, utilizing hydride generator (HG)-multicollector inductively coupled plasma mass spectrometry (MC-ICP-MS, Nu Plasma II and Neptune Plus), we evaluated the method's robustness through achieving high-precision δ¹²³Sb(antimony) values (relative to NIST 3102a) in samples with more complex matrices and low-Sb concentrations. By using element doping (^113/111cadmium) coupled with sample-standard bracketing (SSB), a mass discrimination correction model was presented. The Sb isotope ratios of 30 geological reference materials (GRMs) were reported, in which the values of selected GRMs agreed well with previously published values. The SA method was then used to determine four of these with low-Sb samples. The analytical precision (2 standard deviations: 2SD) was found to mainly depend on the sample fraction (f_spl) in the mixture, with the ideal mixing range recommended to be f_spl ≥ 0.50 for single- and ≥ 0.45 for double SA when expected precision is ≤0.09‰ (2SD). The δ¹²³Sb ratio calculated for basalt BCR-2 (Sb = 0.30 μg g⁻¹) was 0.28 ± 0.09‰, and together with the other four GRMs, it confirmed that accurate and precise Sb isotopes can be analyzed for samples with low-Sb using SA. Validating the measurement of Sb isotope composition in low-Sb samples with complex matrices by SA will expand the application of Sb isotopes in environmental, agricultural, life, Earth, and planetary sciences.