Identification of antigenic variants of the PorA protein of Neisseria meningitidis with specific mAbs (serosubtyping) is used in meningococcal strain characterization and the resultant data has been exploited in the design of novel multivalent vaccines against this important pathogen. The reactivity of the P1.10 serosubtyping mAb MN20F4.17 with eight members of the meningococcal P1.10 variable region (VR) family (prototype P1.10 and variants P1.10a-P1.10g), identified by nucleotide sequence analysis of porA genes, was investigated. Analysis of overlapping synthetic octapeptides by ELISA demonstrated that the peptide sequence, QNQRPTL, present only in the prototype P1.10, was sufficient for binding of the mAb. A linear peptide of 14 amino acids, containing the minimum epitope, inhibited binding of mAb MN20F4.17 to whole cells in a competitive ELISA. This binding was weak compared with a tethered peptide or the native protein. In whole-cell ELISA or dot-blot assays using low concentrations of mAb MN20F4.17 only the prototype P1.10 was detected. However, when higher concentrations of antibody were used the prototype P1.10 was detected, together with variants P1.10a, P1.10c and P1.10e by whole-cell ELISA and P1.10a and P1.10c by the immunoblot technique. The variants P1.10b, P1.10d, P1.10f and P1.10g showed no reactivity with mAb under any of the conditions tested. A survey of the porA genes in serogroup B and C strains revealed that the P1.10a variant, rather than the prototype P1.10, was the most common member of the P1.10 VR family in England and Wales. These data illustrate: (i) the problems associated with epidemiological analyses that rely solely on monoclonal antibodies; (ii) the importance of using defined assay conditions for serosubtyping; and (iii) that genetical analyses provide more reliable information than serological data based on murine reagents for the design of candidate vaccines that include PorA.