A patient who was in therapy slept in a lab and was awakened whenever evidence of EEG and sleep patterns showed that he was dreaming. He was asked to relate his dream with associations, which were obtained also in subsequent therapy hours. In the case of sequential dreams it appeared that the same or a limited number of conflicts were worked upon. The organization of each dream seemed to depend upon the solutions in the previous dreams, that is, toward bolder gratification or greater defensive reactions.