Four herbicide experiments were conducted in a San Antón soil at Juana Díaz Substation from 1985 to 1988. In the first experiment, conducted in 1985-86, we found that post emergence oxyfluorfen (0.14, 0.28 and 0.56 kg ai/ha) preceded by preemergence DCPA (11.2 kg ai/ha) plus hand weeding effectively controlled broadleaf weeds (93-97%) for 9 weeks. Onion yield increased proportionately with the increasing rates of oxyfluorfen. In the second experiment, established in 1986-87, preemergence metolachlor (2.8 and 5.6 kg ai/ha) caused high phytotoxicity on onion and affected its yield. Oxyfluorfen (0.14 kg ai/ha) applied in mixture and sequentially with fluazifop-P (0.24 kg ai/ha) provided an excellent weed control with onion yields similar (P = 0.05) to that obtained in the hand-weeded check. In two subsequent experiments established in 1987-88 we found that two or three applications of either oxyfluorfen (0.14 kg ai/ha) 4- fluazifop-P (0.25 kg ai/ha) or oxyfluorfen (0.14 kg ai/ha) + sethoxydim (0.25 kg ai/ha) mixtures gave excellent grass and broadleaf weed control. However, no significant differences (P = 0.05) in onion yield were detected by increasing the number of herbicide applications in either experiment.