Wilms tumor remains a fascinating model for understanding how genes important in normal human embryogenesis can also contribute, in their mutant form, to cancer development in childhood. The cloning of the first Wilms tumor gene, WT1, in 1989, laid the framework for a model but also emphasized the underlying genetic complexity of this embryonal kidney cancer. Despite longstanding evidence for additional Wilms tumor gene loci, by 1997 very few of these have been cloned and none has yet been proven to be involved in Wilms tumorigenesis in man. However, the potential biological properties of these candidate genes suggest that disregulation of fetal mitogens, such as insulin-like growth factor 2, may be pivotal. Nephrogenesis is clearly sufficiently flexible to absorb many genetic errors. At present, it is unclear how often a simple two mutation model may account for Wilms tumor. Understanding how the various Wilms tumor genes interrelate, if indeed they do, awaits their identification. Piecing together these pathways may eventually lead to logical targets for therapeutic interventions.