The perennial weeds under study i.e. cogongrass "Iniperata cv'indnca (L.) p.", bermudaqrass "Cynodon dactylcn CL) pors.", nutsedge "Cyperus rotundas L." and field bindweed " Convolvulus arvensis L." are cnsidercd to De the dominant noxious weeds in most pa its of the world. Ail treatments gave significant effect on controlling the perennial weeds mixture mentioned above. Glyphosate (Roundup 4b %) at 4.0 L/fed. eitliei used once or twice was the superior herbicide for controlling the perennial weeds .That was true in the four surveys at 3, 6, 9 and .12 weeks and in both seasons, this may be due tc the fact that trns herbicide has a non-selective effect and transiacte effectively in the growing part of annual and perennial weeds. According to the reduction of the peiennial weeds population estimated by visual assessment and the nuniber/nr', the treatments could be arrangeed in the a descending order: RounoTip at 4.0 I/fed, used twice and once, Fusi'dde Super at 2.0 I/ fed usco tw:ce and once, mulching by banana leaves, hand hoeing at four times and the lowest reduction was the cutting at three times in the four surveys in the first season, Roundup al 4.0 I/fed, used twice and once, hand hoeing tour times, Fusilade Super used 2.0 !/ fed used twice, mulching by leaves, Fusi!ade Super at 2.0 I/ fed used once and the iowest was the cutting three times, compared to un treated control in the four surveys in the second season.Another study, was conducted to estimate the predictive equations between the effective dose of glyphosate (Herbazid) and killing percent of the above four perennial weeds. The best control percent was done by using 4 l/fed. Concentration of four weed under study.