Two experiments were conducted to examine whether sensory preconditioning is demonstrable in the evaluative conditioning paradigm. Both experiments included between- and within-subjects control conditions. In Experiment 1, forward pairings of the preexposed stimuli were used, whereas in Experiment 2, forward and backward pairings were employed. The results show that affective–evaluative learning is sensitive to a sensory preconditioning procedure, thus demonstrating that evaluative conditioning does not require the direct experience of a valued event. The subjects did not show any signs of verbal knowledge of the stimulus arrangements. The results are discussed in the context of recent models of classical and evaluative conditioning.