Abstract A small plot experiment was carried out to investigate the effect of forest floor disturbance on establishment of grasses and sedges and to test whether addition of a residual herbicide in tank mix with Roundup could reduce grass and sedge reinvasion. Seven residual herbicides, each at three rates of application, were tested against a Roundup control. Herbicides were Hyvar X (bromacil), Karmex (diuron), Racer (fluorochloridone) (R-40244)), Devrinol (napropamide), Surflan (oryzalin), Oust (sulfometuron methyl), and Sinbar (terbacil). Herbicides were applied on three dates of application at two northwestern Pennsylvania sites. Half of each plot was mechanically disturbed to simulate logging disturbance. Response data were collected for 2 or 3 years after treatment application. Soil disturbance was a significant factor in promoting emergence of grass and sedge; most germination occurred during the growing season immediately following treatment. Little grass or sedge emerged on undisturbed areas. In the second growing season following treatment, plants that did emerge grew to full size. Two of the tank mixes, Roundup plus Hyvar X and Roundup plus Karmex, gave less control of vegetation on the plot at the time of treatment than Roundup alone. Most Round-up-residual herbicide tank mixes reduced emergence and growth of new grass and sedge plants, regardless of application date; however, the amount and duration of control differed among herbicides. Roundup (1 # ai/ac) plus Surflan (2-4 # ai/ac) and Roundup plus Oust (0.09-0.19 # ai/ac) produced the best results reducing grass and sedge ground cover for at least 3 and 2 years, respectively. By use of Roundup-residual herbicide tank mixes and manipulation of cutting techniques, forest land managers can minimize the impact of grass and sedge on the forest regeneration process in Allegheny hardwood stands. North. J. Appl. For. 7:124-129, September 1990.