We investigated source-monitoring errors in qualitative and temporal order judgments. In Experiment 1, a group of participants saw pictures on the first day and imagined items on the third day. The order was reversed in the second group. The result showed that the picture-imagined group misattributed new items to the perceived source whereas the imagined-picture group misattributed the new items to the imagined source. These results were consistent with those obtained by Bink, Marsh and Hicks (1999). In Experiment 2 and 3, we investigated the relationship between qualitative and temporal judgments. Participants were presented with pictures and were asked to imagine items on both the first and third days. They tended to misattribute the source of the new items to the first day rather than to the third day. The results further indicated that qualitative judgments were always correct when temporal order judgments were correct. In contrast, correct qualitative judgments did not reliably predict correct temporal order judgments. The results therefore indicated that qualitative judgments rely heavily on temporal order judgments.