Around the time of the completion of the draft human genome sequence, biologists heatedly debated the number of genes contained in the human genome. In 2003, GeneSweep, an informal gene-count betting pool that began at the 2000 Cold Spring Harbor Laboratory Genome Meeting, announced Lee Rowen of the Institute of Systems Biology in Seattle to be the winner. His bet (25,946) was the lowest and was closest to what many computer programs predicted from the then draft human genome sequence (1). But, in this issue of PNAS, Nozawa, Kawahara, and Nei (2) tell us that asking about the exact number of human genes is meaningless, because different human individuals have different numbers of genes, and it is not uncommon for one person to have 100 more gene copies than another person. They argue that this large variation reflects genomic drift, random changes of gene copy number in evolution.