With growing public concern about environmental quality, farmers must turn to new plant protection alternatives that minimize the use of agrochemicals. Flaming has been practiced for several years as a means of weed control in noncropped areas (railroad, ditches, etc.), but its selectivity toward crops has yet to be defined. Experiments were conducted in the ICG-Propane laboratory at Laval Univ. to determine the temperature needed to kill weeds and the temperature that corn could tolerate. Four weed species were studied: Amaranthus retroflexus, Brassica kaber, Chenopodium album, and Setaria viridis and each species was tested at three growth stages: 0–2, 4–6, and >8 leaves. Corn tolerance was tested at four growth stages: coleoptile, 0–2, 4–6, >8 leaves. All plants were grown in the green-house and were submitted to different combinations of operation speeds and of propane pressures, giving 10 temperature intensities ranging from 110 to 390C. The response of each species was evaluated by measuring its height and dry biomass 2 weeks after treatment. The threshold temperature for corn was below 200C; above this temperature, significant corn injury occurred at all growth stages tested. The corn growth stages most tolerant to heat were coleoptile and >8 leaves. While the most sensitive was 4–6 leaves. All weeds tested were sensitive to heat at 0–2 leaf stage. Amaranthus retroflexus and Chenopodium album were controlled until six leaves with temperatures that were not harmful to corn. Weeds with more than eight leaves needed higher temperature, and control rarely reached 60%. Flaming could be a selective method of weed control if operated at a temperature of 170C. Selectivity can be increased by creating a growth differential between corn and weeds.