The current number of 9422 symbols for human gene names (http://www.gene.ucl.ac.uk/nomenclature/) is expected to increase 7- to 15-fold over the next 2 years. In and around each gene, a tremendous degree of single-nucleotide polymorphism (SNP) heterogeneity is now realized to exist. This review is intended to be visionary, to point out some of the enormously complex nomenclature issues that we face, and to offer some reasonable solutions to these issues. For example, I believe that a 'gene' should be defined as that region from the furthest 5'-ward enhancer to at least 150 bases downstream of the last exon. Just as established rules are critically important for the systematic naming of all new genes, standardized nomenclature rules for the naming of allelic variants are also desperately needed. The evolving consensus for naming the alleles of all human genes (ideally based on evolutionarily diverging haplotype patterns) is described herein. Because of the anticipated explosion in finding new genes and allelic variants due to high-throughput resequencing and DNA-chip technologies, this excess of new knowledge will undoubtedly overwhelm their publication by scientific journals alone. I suggest that the best approach to this staggering 'information overload' is to place the data on appropriate web sites--with numerous links between sites, and frequent updates of all information--so that colleagues in all fields of medical and genetic research can remain knowledgeable. Examples of successful web sites to date include those for the cytochrome P450 (CYP) genes and human CYP alleles, UDP glycosyltransferase (UGT) genes and human alleles, human N-acetylaminotransferase (NAT2, NAT1) alleles, and aldehyde dehydrogenase (ALDH) genes and human alleles. Many more web sites will be necessary. For each site, the webmaster will need to be responsible, accurate, energetic, highly organized, and keen to keep the site current. I believe that interactive discussions on these sites should be encouraged, and advisory committees must be willing to check frequently to ensure that all new information is accurate. Lastly, for the field of molecular epidemiology, the importance of correlating an informative genotype with an unequivocal phenotype is emphasized, and the emerging realization that racial and ethnic groups are highly admixed is summarized and updated.