||Canebrakes are dense stands of Arundinaria gigantea that once were prominent in the southeastern United States. Rivercane still occurs as an understory component of bottomland hardwood forests, but with intense agricultural development over the past 200 years, canebrakes are now considered a critically endangered ecosystem. There is increasing interest in the use of A. gigantea in riparian restoration and soil stabilization. Additionally, A. gigantea possesses cultural significance to Native Americans as a major component of construction, basketry, and weapon making. This research assesses response of A. gigantea to existing plant assemblages and to site preparation techniques, when planted with primarily native grasses (big bluestem, Andropogon gerardii, and indiangrass, Sorghastrum nutans), versus primarily exotic grasses (johnsongrass, Sorghum halepense, and bermudagrass, Cynodon dactylon). To assess site preparation techniques, soil tillage and herbicide application (a combination of 2.0% active ingredient glyphosate followed by application of a pendimethalin-based pre-emergent) were applied in a factorial design to plots dominated by the above‐mentioned plant assemblages. Effectiveness of the treatments and competitiveness of A. gigantea with existing vegetation were determined from measurements of shoot height and diameter, production of new shoots, and a plant size index (approximating canopy volume). Analyses of the first year’s data indicated that A. gigantea mean plant size was influenced by herbicide site prep, as well as soil tillage, when planted in exotic vs. native grass plots. Although soil tillage was not a significant factor by itself, A. gigantea mean plant size appeared to respond differently in the native vs. exotic plots. Tilling in the exotic plots caused a doubling of mean plant size (13,129 cm3) compared to untilled treatments (5151 cm3). Similarly, herbicide also was a non-significant factor by itself, but A. gigantea mean plant size appeared to respond differently in the native vs. exotic plots. Herbicide application in the native plots caused a significant 95% decrease in mean plant size (462 cm3) compared to no‐herbicide treatments (9634 cm3). Mean shoot height and the number of shoots were significantly higher in the native grass plots vs. the exotic grass plots. From these first-year data, it can be concluded that A. gigantea planted into plots dominated by non-native plants benefited significantly more from site preparation (tilling, herbicide) than cane planted into native-species‐dominated assemblages. However, it is not yet know whether these patterns have persisted through year two. If the patterns observed in year one persist, this research should contribute to improving the success of future cane restoration projects by indicating when intensive site preparation is necessary, and when it should be avoided.