Free radical grafting of glycidyl methacrylate (GMA) onto molten polypropylene (PP) was studied in a co-rotating twin screw extruder. Grafting yields of GMA obtained under various experimental conditions along the screw length allowed for a good appreciation of the effects of chemical parameters (the presence of styrene and the concentrations of peroxide and monomers) and those of processing parameters (feed rate, screw speed, and specific throughput). Similar to the results obtained in a batch mixer,(1) free radical grafting of GMA carried out in the extruder in the presence or absence of styrene proceeded rapidly, as it was virtually completed half-way down stream of the extruder. Additionally, the presence of styrene as a second monomer increased the GMA grafting yield greatly with reduced PP chain degradation. The ultimate GMA grafting yield increased with increasing concentration of peroxide, 1,3-bis(tert-butylperoxyisopropyl)benzene. This similarity between the batch mixer and the extruder is related to the fact that in both cases it is the concentration of the peroxide and its half lifetime that determine the grafting rate and the ultimate grafting yield. On the other hand, the GMA grafting yield decreased with increasing screw speed or feed rate. For a particular specific throughput (the ratio of throughput to screw speed), an increase in throughput with a concomitant increase in screw speed brought about a decrease in GMA grafting yield. It was concluded that the GMA grafting yield is affected primarily by the residence time in the zone in which free radicals are not depleted. The effects of screw speed, feed rate, and specific throughput manifest mainly through this local residence time distribution. Specific energy is not a good measure of the performance of the extruder with respect to the free radical grafting of GMA onto PP.