Measurements are reported on the perceived orientation of a line segment in a frontal plane (test line). For an isolated test line, perceived orientation is biased towards the horizontal H or vertical V, whichever is nearer, H and V are named preference orientations. The presence of a second line (induction line) brings about an extra bias towards the perpendicular of the induction line: induced preference orientation. The phenomenon is well known from many visual illusions. H or V induction lines exert greater influences than do oblique induction lines, H or V test lines are less susceptible to induction effects than are oblique test lines. If a short test line is replaced by two dots at its ends, the induction effect remains unchanged. If the test line is presented to one eye and the induction line to the other, the induction effect remains almost unchanged, which is indicative of extensive cooperation between the two eyes. In a qualitative model, in which preference orientations are represented by a maximum sensitivity of orientation sensors along the slant axis, the induction effect can be incorporated by an additional assumption. Physiological evidence as to the origin of the induction effect is scanty.