The affect-as-information hypothesis states that negative mood results in a more analytic and positive mood in a more heuristic processing style. Evidence comes from a wide range of research areas, e.g., on memory or person perception. Studies on implicit learning, however, are scarce and evidence mixed. Therefore, the present study examined the influence of mood on implicit sequence learning and the formation of explicit knowledge. An incidental sequence learning task was used during which negative, neutral, and positive background pictures were used to induce mood states in three groups of participants. Consistent with the affect-as-information hypothesis, the results show that a less positive mood resulted in more explicit sequence knowledge. In contrast, positive mood did not enhance implicit learning. It is speculated that implicit regularity learning as an automatic by-product of task processing might be less susceptible to affective processing biases than other, more complex types of cognitive processing