In order to determine the best time to use and the adequate dose of four herbicides to control weeds in dryland chickpea (Cicer arietinum L.) fields, we performed the present experiment in 4 x 5 m plots. Fourteen treatments were carried out that from 1 to 9 included trifluralin. Treatments 1, 2, and 3 were with increasing doses of trifluralin (480, 720, and 960 g ai ha-1) applied 30 days before planting. Treatments 4, 5, and 6 included increasing doses of trifluralin (480, 720, and 960 g ai ha-1) applied 15 days before planting. Treatments 7, 8 and 9 consisted of increasing doses of trifluralin (480, 720, and 960 g ai ha-1) applied at the time of planting. Treatments 10, 11, and 12 included pyroxasulfone (85 g ai ha-1), flumioxazin (51 g ai ha-1) and imazethapyr (100 g ai ha-1), respectively. These last three treatments were carried out at the time of planting; treatments 13 and 14 were: weed-infested (without weed control) and weed-free (manual weeding during the entire season). Flumioxazin 66% and pyroxasulfone 57% (mean of two samples) reduced weed dry weight compared to uncontrolled treatment. The results showed that the treatments were significantly different for 100-seed weight, biological yield, and seed yield of chickpea. Weed-infested and weed-free plants had the lowest and highest grain yield. Herbicide treatments of flumioxazin, trifluralin 960 g ai ha-1, and pyroxasulfone at planting produced 55%, 44%, and 40% higher grain yield than the weed-infested plots. Also, none of the herbicide treatments reduced chickpea yield and biomass. The herbicide residues had no adverse effect on wheat growth in the next crop season.