Chicks were trained to discriminate (1) between two boxes of the same colour on the basis of their positions; (2) between two boxes of different colours on the basis of their positions or on the basis of their colours. No significant differences between these tasks appeared. Addition of colour and position improved learning when the boxes were entirely coloured but not when differently coloured discs were placed over the lids of two identical boxes. It is argued that in simultaneous discrimination learning chicks rely primarily on the “relative” position of the goal objects (i. e. in relation to each other) and utilize featural information (colour) to segregate the goal objects as distinct perceptual units. Results also stressed the role of “object characteristics” as opposed to the traditional notion of “cues”.