The nightly amount of dreaming sleep of 6 male volunteer college students was drastically reduced for 3 consecutive nights by each of 2 methods, dream interruption and partial sleep deprivation. Psychological reaction to the reduced dreaming was studied with 3 types of data: tape-recorded and transcribed dream reports; field notes maintained throughout the experiment on the subjects'' behaviors, conversations, preoccupations and responses to queries about the daily rounds and moods; and psychometric tests including the D scale of the MMPI [Minnesota Multiphasic Personality Inventory], digit span forward and backward, serial subtraction and a word association test. Limited psychological testing failed to disclose any significant changes with deprivation. Behavioral observations, however, were for the most part consistent with early reports by Dement and by Fisher. Some subjects developed intense hunger and special food cravings; all reported some increase in appetite during 1 or both deprivation series. There was, however, no increase in oral responses to a word association list and in contrast to Fisher''s report no increase in oral imagery in the REM [rapid eye movement] dream fragments obtained on dream interruption nights. Some instances of disturbances in subjects'' relationship to reality or feelings of reality were observed and reported. The intensity of reaction to the experimental deprivations was notable, with instances of childish behavior, the emergence of prominent masochistic and even paranoid themes, some injuries to self, and 1 instance of a subject tearing off the well secured electrodes during sleep. There was a statistically significant increase in aggressive content in REM dream fragments as deprivation increased. In the absence of additional control studies, findings cannot definitively be attributed to a reduction in dreaming per se, or to a reduction in REM sleep. It is possible that the disturbances were reactions to the experimental situation, although the intensity of response may have been magnified by the lack of opportunity to discharge or master the excitation in dreaming. There was also a statistically significant increase in the amount of manifest distortion in interrupted REM dreams as deprivation increased. It was tentatively speculated that this change could represent some kind of acceleration of the dream process or a compensatory intensification of dreaming with increased psychic discharge value. Although this finding is provisional, it is of particular interest because it has a closer, more direct bearing on a "need to dream" than other kinds of evidence reported here or in prior studies.