Host-plant preferences were assessed by field measurement of grasshopper abundance in relation to measurements of floral community and by laboratory host-plant preferences. Correlations between grasshoppers and plants (biomass and percentage cover) were based on data gathered at 29 study sites in five distinct Florida habitats (including disturbed, freshwater marsh, high pine, oak hammock, and swamp). Relationships among plants in these habitats and 10 abundant grasshopper species were examined to clarify differences in grasshopper assemblages among habitats. The grasshoppers studied were Aptenopedes sphenarioides Scudder, Chortophaga australior (Rehn and Hebard), Eritettix obscurus (Scudder), Melanoplus bispinosus Scudder, Melanoplus querneus Rehn and Hebard, Paroxya clavuliger (Serville), Schistocerca americana (Drury), Schistocerca ceratiola Hubbell and Walker, and Spharagemon crepitans (Saussure). In addition to correlations among individual plants and the abundance of grasshoppers, multiple regression analysis was used to assess how groups of plants were related to grasshopper abundance. These analyses identified grasshopper-plant associations, but some plants were found to be host plants, whereas others were indicators of preferred habitats (indicator plants). Host-plant preferences were determined in the laboratory using five plant-choice tests to help distinguish between host and indicator plants. The 10 grasshopper species examined for laboratory host-plant preferences were the same species scrutinized in the field study. In almost every case, grasshoppers showed specific plant preferences, and the preference studies successfully distinguished between host plants and indicator species. Similar results were obtained whether biomass or percent cover measurements were used to assess the floral community, although percent cover is much easier and faster to determine.