Evolutionary biologists have always been interested in the origin and evolution of new genes. The most obvious mechanisms of their formation are various types of chromosomal and intergenic rearrangements, suggesting the use of existing genes as the source material. The possibility of the origin of a fully functional gene from noncoding DNA, i.e., de novo, was not rejected but it was practically considered as impossible until recently. Nevertheless, in 1996, after analysis of the yeast Saccharomyces cerevisiae genome, the first experimental evidence of the possibility of de novo gene formation was obtained. Ten years later, genes that do not have homologs, presumably having originated de novo, were found in Drosophila. The relatively high probability of the occurrence of genes de novo, estimated in bioinformatics studies, raised the interest in this topic and made the search for genes of this type relevant. Recently, the number of works devoted to the problem of the emergence of de novo genes in different organisms, including humans, is constantly growing, demystifying this phenomenon. Nevertheless, there are still many questions that require theoretical and practical research. This review is devoted to the problem of finding and characterizing genes that have arisen de novo, as well as to the proposed mechanisms of their occurrence.