Despite the increasing popularity of multilocus sequence typing (MLST), the most appropriate method for characterizing bacterial variation and facilitating epidemiological investigations remains a matter of debate. Here, we propose that different typing schemes should be compared on the basis of their power to infer clonal relationships and investigate the utility of sequence data for genealogical reconstruction by exploiting new statistical tools and data from 20 housekeeping loci for 93 isolates of the bacterial pathogen Neisseria meningitidis. Our analysis demonstrated that all but one of the hyperinvasive isolates established by multilocus enzyme electrophoresis and MLST were grouped into one of six genealogical lineages, each of which contained substantial variation. Due to the confounding effect of recombination, evolutionary relationships among these lineages remained unclear, even using 20 loci. Analyses of the seven loci in the standard MLST scheme using the same methods reproduced this classification, but were unable to support finer inferences concerning the relationships between the members within each complex.