Mercury (Hg) isotopic analysis via multi-collector inductively coupled plasma (ICP)-mass spectrometry (MC-ICP-MS) can provide relevant biogeochemical information by revealing sources, pathways, and sinks of this highly toxic metal. In this work, the capabilities and limitations of two different sample introduction systems, based on pneumatic nebulization (PN) and cold vapor generation (CVG), respectively, were evaluated in the context of Hg isotopic analysis via MC-ICP-MS. The effect of (i) instrument settings and acquisition parameters, (ii) concentration of analyte element (Hg), and internal standard (Tl)-used for mass discrimination correction purposes-and (iii) different mass bias correction approaches on the accuracy and precision of Hg isotope ratio results was evaluated. The extent and stability of mass bias were assessed in a long-term study (18 months, n = 250), demonstrating a precision ≤0.006% relative standard deviation (RSD). CVG-MC-ICP-MS showed an approximately 20-fold enhancement in Hg signal intensity compared with PN-MC-ICP-MS. For CVG-MC-ICP-MS, the mass bias induced by instrumental mass discrimination was accurately corrected for by using either external correction in a sample-standard bracketing approach (SSB) or double correction, consisting of the use of Tl as internal standard in a revised version of the Russell law (Baxter approach), followed by SSB. Concomitant matrix elements did not affect CVG-ICP-MS results. Neither with PN, nor with CVG, any evidence for mass-independent discrimination effects in the instrument was observed within the experimental precision obtained. CVG-MC-ICP-MS was finally used for Hg isotopic analysis of reference materials (RMs) of relevant environmental origin. The isotopic composition of Hg in RMs of marine biological origin testified of mass-independent fractionation that affected the odd-numbered Hg isotopes. While older RMs were used for validation purposes, novel Hg isotopic data are provided for the latest generations of some biological RMs.