In Egypt, alfalfa represts a suitable choice for forage cultivation expansion, since, the available land and water are oflower quality. The early seeding stage of alfalfa is the most val near able to weed competition. Research results regardingimprovement of alfalfa tolerance to glyphosate in Egypt is relatively scare. The recent study was an attempt to tracevariability in glyphosate tolerance of alfalfa germplasm. The improvement in tolerance due to recurrent selection was alsoconsidered. Alfalfa plant materials (Medicago sativa,L. ) used in that recent study were five base populations .Two cyclesof recurrent selection for Glyphosate tolerance were imposed on each base population. Evaluation of selected cycles (C1and C2) along with base populations (C0) was carried-out for each population as a split plot design with Glyphosatetreatment (+ and -) as main plots and populations (C0, C1 and C2) as a sub –plots. Glyphosate treated plots were evaluatedfor injury characters, i.e.; total chlorophyll, injury level, percent of death and shikimic acid content.Significant (p≥0.01) variations were detected among the studied alfalfa population. Also, the recorded values for allinjury indicator characters significantly varied among selection cycles. In the meantime, the interactions between thestudied population and selection cycles were significant (p ≥0.05). Chlorophyll content expressed as units was notchanged after one cycle of selection (30.22 and 29.15 spade for base population (C0) and first cycle of selection (C1) as anaverage over the studied population, respectively). While, the second cycle of selection resulted in significantly higherlevel of total chlorophyll content (34.51 spade). This might indicates that tolerance to glyphosate was associated to higherchlorophyll content. The value of injury level decreased in both of Siwa and Hasawi populations after the first cycle ofselection to glyphosate tolerance(-19.51 and -13.22%, respectively). Meanwhile, the progress of selection for glyphosatetolerance in alfalfa populations, gave a substantial decrease in value of injury level in all studied populations, except for,Siwa population that recorded an increase in value of injury level (-47.05, -44.44, -40.91 and -4.35 percent of therespective value recorded for injury level at the first cycle of selection in C.U.F101, Baladi 1, Sirivar and Hasawipopulations, respectively).Change in shikimic acid content represented by a decrease reached -35.05, -22.99, -5.49 and -4.26% in Siwa, Baldi 1, C.U.F 101 and Sirivar populations, respectively, after the first cycle of selection for glyphosatetolerance.Only Hasawi population recorded an increase in shikimic acid content reached +251.8% relative to base population.While, after the second cycle of selection an increase in shikimic acid content reached +63.60 and +4.688% for C.U.F101and Baldi1 populations relative to (C1). Sharing genes among tolerant individuals, increase the frequency of genesresponsible for tolerance, consequently expressing higher levels of cide- tolerance. The recent results assume thepossibility of obtaining glyphosate tolerant alfalfa population depending on frequent cycles of recurrent selection.