A study was conducted to evaluate differences in rangeland grasshopper communities over environmental gradients in Gallatin Valley, Montana, USA. The concept of habitat type (Daubenmire 1966) was used as a basis for discriminating between groupings of patches based on vegetation. A total of 39 patches were selected that represented five recognized grassland habitat types (Mueggler and Stewart 1980), as well as two disturbed types (replanting within a known habitat type). Repeated sampling in 1988 of both the insect and plant communities yielded a total of 40 grasshopper (19 664 individuals) and 97 plant species. Detrended Correspondence Analysis (DCA) indicated that patch classifications based on presence and percent cover of plants were appropriate and showed good between-group (habitat type) separation for patches along gradients of precipitation/elevation and plant community complexity. Results from undisturbed habitats showed that plant and grasshopper species composition changed over observed environmental gradients and suggested that habitat type influenced not only species presence, but also relative abundance. Discussion is presented that relates results with patch-use and core and satellite species paradigms.