We have previously reported significant differences in the extent of somatic point mutation seen in immunoglobulin variable region genes associated with different IgG constant region genes. The mean mutation counts seen corresponded to the position of the constant region genes within the locus. IGHV genes associated with the most 5’ C gene (encoding IgG3) had the least mutations. The study population for this analysis was a rural community in Papua New Guinea, where parasitic infections are endemic. We have now repeated our study using 454 pyrosequencing to generate IGHV sequences from 8 healthy, urban residents of Sydney, Australia. The same relationship was seen between mutation counts and IgG subclass. The mean mutations counts were: IgG3 16.5, IgG1 18.6, IgG2, 19.4 and IgG4 21.8. These differences were highly significant (IgG1 vs IgG2: p<0.01; All others p<0.0001). By analyzing the distribution of replacement mutations between the CDRs and Framework regions of the IGHV genes, the extent of antigen selection could be measured for the different subclasses. Selection was significantly stronger in IgG2 sequences, despite the association of this isotype in the literature with T independent responses. This result also confirms findings from our PNG study. Mutations in IGHV genes associated with the IgA subclass-encoding genes were also analyzed in Australian residents. While the mean numbers of mutations in IgA1 (19.9) and IgA2 sequences (19.7) were similar, IgA2 shows significantly more evidence of antigen selection than IgA1 sequences. These results suggest that B cells follow a programmed pathway of sequential switches between each of the IgG subclass-encoding IGHG genes, and that there are predictable differences in the relative antigen-binding affinities of the different IgG subclasses. These inferred affinities underpin our recently-proposed Temporal Model of human IgG function, which seeks to explain how the four IgG subclasses work together to deliver protective immunity.