Structured demographic models can be valuable for assessing the status of at-risk plant species and for guiding efforts to protect and manage them, but they may be infrequently employed in assessments of species of conservation concern because they require data that are not often available to those charged with management. We constructed and analyzed matrix projection models for three populations of the threatened perennial plant species Euphorbia telephioides to reevaluate a prior assessment of stable status and to compare the conclusions of models based on different amounts of data. In contrast to the prior assessment, all of our models predicted population decline for all three populations. Projected growth rates were similar for models based on sparse sampling and those based on a larger more representative sample of individuals, but the models supported contrasting patterns of the sensitivity of population growth rate to individual demographic transitions. Models based on more complete data supported greater elasticity of growth rate to the demographic fates of small non-flowering plants, which had among the lowest elasticity values for models based on more sparse data. We conclude that even models based on limited data can be useful to species status assessments, but that more complete data may be necessary to identify targets for effective management.