The human and mouse genomes contain 21 and 20 connexin genes, respectively. During the last 10-year period, genetic research on connexins has been stimulated by two parallel approaches: first, the characterization of genetic diseases that are caused by connexin mutations and, second, the generation and characterization of connexin knockout (null) mutated mice in which the coding region of nearly all connexin genes has been deleted. We summarize the current results of each of these two approaches. More recently, first results have been published in which connexin point mutations in human connexin genes were inserted at the corresponding position of the orthologous mouse gene. Under these conditions, the mutated connexin protein is expressed, in contrast to a connexin null mutation, and its interaction with other connexin isoforms or other connexin-binding proteins can be maintained. In this review, we discuss advantages and problems of such an approach and possible implications regarding the mechanism of the disease. The long-term goal is to understand the biologic function of each connexin isoform and the contribution of these proteins to the physiology of the corresponding organs in health and disease.