Herbivores may negatively impact plants by reducing biomass and decreasing fecundity. Rare or endangered plants may be particularly vulnerable to herbivore damage. However, plants may compensate for herbivory by increased growth or reproductive output. We investigated compensatory growth and reproductive effort in Liatris ohlingerae (S.F.Blake) B.L.Robe, a federally-listed herb narrowly endemic to pyrogenic Florida scrub on the Lake Wales Ridge, by conducting biweekly surveys of observed herbivory and an herbivore exclusion experiment. In the field surveys, we assessed the effects of vertebrate herbivory (“topping”) and invertebrate leaf herbivory on plant size and fecundity and compared observed levels of herbivory across sites with different disturbance histories (scrub vs. roadside; scrub sites with differing time-since-fire). In the herbivore exclusion experiment, we compared the effects on plant size and fecundity of excluding vertebrate herbivores (toppers) only vs. excluding both vertebrate and invertebrate herbivores. We found that 62% of plants in the survey had one or more topped stems and that topping occurred more frequently in recently burned sites (< 8 years since fire vs. > 14 years since fire) and in scrub than in roadside sites. Topped plants were significantly shorter and had a greater number of stems (untopped plants only produced 1 stem). Topped plants were less likely to flower, began flowering later in the growing season, and produced fewer inflorescences, infructescences, achenes, and full achenes than untopped plants. Fifty-two percent of surveyed L. ohlingerae plants also showed evidence of invertebrate leaf herbivory with a mean of 24% of leaves damaged. Plants damaged by invertebrate herbivores did not differ significantly in growth or reproduction compared with undamaged plants. In the experiment, exclusion of vertebrate herbivores resulted in taller plants, fewer stems, higher flowering frequency, earlier first flowering, and greater production of inflorescences, infructescences, achenes, and full achenes per plant; plants gained no additional benefits when invertebrates were excluded. Our study demonstrates that L. ohlingerae partially compensates for vertebrate herbivore damage at some sites, depending on the disturbance history. There are population level consequences of undercompensation; vertebrate topping resulted in a 30% reduction in mean fecundity. Future research is necessary to determine if undercompensation jeopardizes the long-term persistence of L. ohlingerae populations.