Habits are underexplored in research on evidence-based psychological treatments (EBPTs). We recruited participants (N = 286) with sleep problems via Mechanical Turk for an experiment to shift wake-up habits, a key target of EBPTs for sleep problems. Participants were randomly assigned to a control (i.e., psychoeducation about healthy wake-up habits) or one of five active habit-change strategies: substitution with RISE UP, awareness training, vigilant monitoring, implementation intentions, and values. New and old wake-up habit strength, sleep disruption, and sleep-related impairment were assessed at baseline, six-week follow-up, and three-month follow-up. Aim 1 tested within- and between-condition change in the outcomes. Aim 2 tested whether change in wake-up habit strength predicted improvements in sleep disruption and sleep-related impairment. Except the values condition, all habit-change strategies and the control were significantly associated with within-condition improvements at 6-week follow-up and 3-month follow-up in: new habit strength (d = 0.81 to 1.68), old habit strength (d = -0.63 to -1.04), sleep disruption (d = -0.97 to -1.98), and sleep-related impairment (d = -0.60 to -1.65). Few differences between conditions emerged. Across conditions, more than 50% of participants met thresholds for clinically meaningful improvement, except the values condition at 3-month follow-up. Change in new and old habit strength significantly predicted change in sleep problems. Key limitations included: an exclusively online study design, dropout rate, and sample collected via Mechanical Turk using self-report measures without formal assessment of sleep diagnoses. Future research should investigate the clinical presentations and EBPT skills for which these habit-change strategies are most effective.