A total of 36 toddlers and preschool children exhibiting bedtime tantrum activity were randomly assigned to one of three groups: positive routines, graduated extinction, or control. Positive routines involved changing the child's bedtime to coincide with when he naturally fell asleep, as well as parent and child engaging in a series of four to seven enjoyable activities before the child being placed in bed. During the treatment period, bedtimes were systematically scheduled earlier such that the child went to bed at the time parents had originally used. Graduated extinction consisted of the parent putting the child to bed and ignoring the tantrum activity for increasingly longer amounts of time throughout the treatment. Children in these two treatment groups had tantrums less frequently and for shorter periods than control subjects during 6 weeks of treatment and during two follow-up observations 3 and 6 weeks after treatment. Although both treatments were more effective than waiting for the child to outgrow this problem, parents of the positive routine group reported significantly improved marital satisfaction, suggesting additional benefits of this treatment strategy.