Little is known concerning the time-course and structural dynamics of reactivation of compensatory myogenesis in denervated muscle, its initiating cellular mechanisms, and the relationship between this process and the progression of postdenervation atrophy. The purpose of this study was to investigate the interrelations between temporal and spatial patterns of the myogenic response in denervated muscle and progressive atrophy of muscle fibers. Another objective was to study whether reactivation of myogenesis correlates with destabilization of the differentiated state and death of denervated muscle cells. It has remained unclear whether muscle fiber atrophy was the primary factor activating the myogenic response, what levels of cellular atrophy were associated with its activation, and whether the initiation and intensity of myogenesis depended on the local and individual heterogeneity of atrophic changes among fibers. For this reason, our objective was also to identify the levels of atrophic and degenerative changes in denervated muscle fibers that are correlated with activation of the myogenic response. We found that the reactivation of myogenesis in the tibialis anterior and extensor digitorum longus muscles of the rat starts between days 10-21 following nerve transection, before atrophy has attained advanced level, long before dead cells are found in the tissue. Formation of new muscle fibers reaches its maximum between 2 and 4 months following denervation and gradually decreases with progressive postdenervation atrophy. The myogenic response is biphasic and includes two distinct processes. The first process resembles the formation of secondary and tertiary generations of myotubes during normal muscle development and dominates during the first 2 months of denervation. During this period, activated satellite cells form new myotubes on live differentiated muscle fibers. Most of the daughter myotubes in 1- and 2-month denervated muscle develop on the surface of fast type parent muscle fibers, and some of the newly formed muscle fibers express slow myosin. Some fast type parent fibers are weakly or, more rarely, moderately immunopositive for embryonic isomyosin. This indicates that reactivation of myogenesis may also depend on the fiber type. The level of atrophy, destabilization of the differentiated myofiber phenotype, and degenerative changes of individual fibers in denervated muscle are very heterogeneous. The myogenic response of the first type is associated predominantly with fibers of average and higher than average levels of atrophy. Muscle cells that undergo a lesser degree of atrophy also form daughter fibers, although with a lower incidence. We did not find any correlation between the size of newly formed fibers and the level of atrophy of parent fibers. The topographical distribution of new myotubes both in the peripheral and central areas of the mid-belly equatorial sections at the early stages following nerve transection indicates that myogenesis of the first type represents a systemic reaction of muscle to the loss of neural control. These data indicate that activation of the myogenic response does not depend on cell death and degenerative processes per se. The second type of myogenesis is a typical regenerative reaction that occurs mainly within the spaces surrounded by the basal laminae of dead muscle fibers. Myocytes of different sizes are susceptible to degeneration and death, which indicates that cell death in denervated muscle does not correlate with levels of muscle cell atrophy. The regenerative process frequently results in development of abnormal muscle cells that branch or form small clusters. Replacement of lost fibers becomes activated between 2 and 4 months following nerve transection, i.e., mainly at advanced stages of postdenervation atrophy, when cell death becomes a contributing factor of the atrophic process. In long-term denervated muscle, the first and second types of myogenesis occur concurrently, and the topographical distribution of the myogenic response becomes more heterogeneous than during the first weeks following denervation. Thus, our data demonstrate differential temporal and spatial expression of two patterns of myogenesis in denervated muscle that appear to be controlled by different regulatory mechanisms during the postdenervation period.