AbstractThe specificity of eukaryotic DNA organization into loops fixed to the nuclear matrix/chromosomal scaffold has been studied for more than fifteen years. The results and conclusions of different authors remain, however, controversial. Recently, we have elaborated a new approach to the study of chromosomal DNA loops. Instead of characterizing loop basements (nuclear matrix DNA), we have concentrated our efforts on the characterization of individual loops after their excision by DNA topoisomerase II‐mediated DNA cleavage at matrix attachment sites. In this review the results of applying this mapping approach are compared with the results and conclusions from studies of nuclear matrix DNA. An attempt is also made to reconsider all data about the specificity of DNA interactions with the nuclear matrix and to suggest a model of spatial organization of the eukaryotic genome which resolves apparent contradictions between these data.