The relationships between several variables and amount of dreaming, defined as time spent in the rapid eye movement stage of sleep, were investigated. Twenty Ss slept in the laboratory for four nights each in accord with a counterbalanced design. Monetary incentive to control amount of dreaming had a small but significant effect on amount of dreaming. On the first experimental night, Ss dreamed less than on succeeding nights; this decrease in dreaming was associated with more spontaneous awakenings on the first night than on succeeding nights. There were stable individual differences in amount of dreaming. The longer the sleep period prior to the first dream of the night, the longer was the first dream.