Quaking aspen (Populus tremuloides Michx.) is a long-lived clonal species in which many genetically identical stems (ramets) arise from a common root system. Establishment by seed is extremely rare in the Rocky Mountain region, where most clones that exist today are thought to have established hundreds or thousands of years ago. However, recruitment of new genetic individuals through sexual reproduction has occurred periodically throughout the Holocene, and widespread establishment of seedling aspen occurred in Yellowstone National Park, USA, following the extensive 1988 fires. We combined extensive survey methods with manipulative experiments to investigate the patterns and mechanisms of new aspen genet establishment, growth, and survival during their first decade of development. Excavation and aging of 173 aspen stems in 1996 demonstrated that 65% had established within the first three years after the 1988 fires, and that none pre-dated 1988. Random amplified polymorphic DNA (RAPD) genetic analyses revealed that 92% of the plants were genetically distinct individuals, and 8% were ramets. Annual surveys of 22 permanently marked aspen seedling plots revealed that 24% of 417 seedlings tallied in 1996 had died by 2000. However, mortality varied greatly among the 22 plots, from <10% to >40%, with greater mortality at lower elevations and where soil organic matter was low. To evaluate the mechanisms underlying seedling persistence or mortality, we constructed exclosures in three sites to protect aspen seedlings from ungulate browsing, and applied four experimental treatments from 1996 to 1998: (1) clipping of current year's growth to simulate browsing, (2) removal of potential competitors including lodgepole pine (Pinus contorta var. latifolia) saplings and herbaceous plants, (3) clipping and competitor removal, and (4) control, i.e., no treatment except protection from browsing. Clipping prevented stem elongation, but removing competitors had no significant effect on height growth of aspen seedlings. Even with protection from browsing, most aspen stems grew slowly (mean increment <25 cm from 1996 to 2000), and many died from causes unrelated to herbivory. Nevertheless, some individuals were >2 m tall in 2003 and appeared vigorous. The aspen cohort that germinated after the 1988 fires appears to be in the earliest stage of a long-term population process, a process that likely will entail a shift from many genetically distinct individuals but few ramets, to relatively few genets having numerous ramets. We predict that many or most of the post-1988 aspen seedlings will die within the next few decades, with little lasting effect on broad-scale vegetation patterns or postfire successional trajectories. However, some new genets appear likely to survive and to establish new aspen clones, with potentially important consequences for demographic and genetic structure of the Yellowstone aspen population.