Seeds of chia (Salvia hispanica L.) are known worldwide for the benefits in human nutrition. However, from the agricultural point of view, weed interference is one of the main limitations in this crop, and restricted basic information about the management with herbicides. The aim of this work was to evaluate the selectivity of herbicides applied in pre-emergence of chia crop (S. hispanica). It was carried out an experiment in the field at the Universidade Estadual do Centro-Oeste, Guarapuava - PR, using randomized complete block design, with ten treatments and four replications. The treatments applied in the pre-emergence of chia were constituted by the herbicides: sulfentrazone (100, 200, 400 and 800 g ha(-1)), oxyfluorfen (720 g ha(-1)), s-metolachlor (1920 g ha(-1)), flumioxazin (50 g ha(-1)), diclosulan (25 g ha(-1)), trifluralin (1800 g ha(-1)) and a control without application. Oxyfluorfen, s-metolachlor, flumioxazin and diclosulan were not selective for application in chia crop in pre-emergence, unlike trifluralin, which showed high selectivity. Sulfentrazone presented viability only at the dose of 100 g ha(-1).