Memory-based ratings of chemosensory mixtures produce response patterns similar to those observed perceptually. This finding is extended in two experiments. In the first, subjects mad judgments of sweetness, sourness and intensity in memory or with perceptually present combinations of sucrose and citric acid. Performance in both conditions was equivalent. Debriefing revealed some explicit knowledge about the way such mixtures interact. A second study investigated the relationship between such explicit knowledge and performance on the memory task. Here, subjects made the same ratings, but sampled, on separate days; (1) capsaicin alone, (2) a flavoured, sucrose, critic acid solution and imagined adding capsaicin, (3) the solution alone and (4) the solution actually with capsaicin. Performance was equivalent across the semi-mental and real mixture. Only sweetness was suppressed. Debriefing revealed that subjects thought all tastes and flavours were affected by capsaicin. Apart from extending previous findings, these studies suggest a role for both explicit and implicit memory processes in mental mixture judgments.