We modeled the population viability of the endangered clonal shrub Florida ziziphus (Ziziphus celata) based on nine years of data from two populations to identify management priorities.  Modeling a critically rare species, such as Florida ziziphus, requires additional consideration for both small sample sizes in parameterization and short-term consequences for persistence in population dynamics.  We compare projection models within these uniclonal populations based on hypothetically independent “plants”.  Seedling recruitment is unknown in the study populations because of cross-incompatibility and is not modeled. The populations modeled here have stable demographies characterized by high survival and stasis, but variable levels of new plant production (clonal recruitment). Population growth rates suggest protracted long-term declines in population size, but predicted extinction rates over 50 years greatly increased when one favorable year is removed from simulations. Changes in new plant production and survival had greater effects on population growth rates than did changes in growth rates of individual plants. Augmenting clonal plant production and protecting survival of new “plants” are short-term management goals.  The establishment of sexually reproducing populations using augmentations and introductions is a long-term necessity for the survival of Florida ziziphus.