In this paper, two different methods are for the first time systematically compared for the determination of selenium in human serum selenoalbumin (SeAlb). Firstly, SeAlb was enzymatically hydrolyzed and the resulting selenomethionine (SeMet) was quantified using species-specific isotope dilution (SSID) with reversed phase-HPLC (RP-HPLC) hyphenated to (collision/reaction cell) inductively coupled plasma-quadrupole mass spectrometry (CRC ICP-QMS). In order to assess the enzymatic hydrolysis yield, SeAlb was determined as an intact protein by affinity-HPLC (AF-HPLC) coupled to CRC ICP-QMS. Using this approach, glutathione peroxidase (GPx) and selenoprotein P (SelP) (the two selenoproteins present in serum) were also determined within the same chromatographic run. The levels of selenium associated with SeAlb in three serum materials, namely BCR-637, Seronorm level 1 and Seronorm level 2, obtained using both methods were in a good agreement. Verification of the absence of free SeMet, which interferes with the SeAlb determination (down to the amino acid level), in such materials was addressed by analyzing the fraction of GPx, partially purified by AF-HPLC, using RP-HPLC (GPx only) and size exclusion-HPLC (SE-HPLC) coupled to CRC ICP-QMS. The latter methodology was also used for the investigation of the presence of selenium species other than the selenoproteins in the (AF-HPLC) SelP and SeAlb fractions; the same selenium peaks were detected in both control and BCR-637 serum with a difference in age of ca. 12 years. It is also for the first time that the concentrations of selenium associated with SeAlb, GPx and SelP species in such commercially available serums (only certified or having indicative levels of total selenium content) are reported. Such indicative values can be used for reference purposes in future validation of speciation methods for selenium in human serum and/or inter-laboratory comparisons.