The aim of this review is to evaluate the effectiveness of cognitive behavioral therapy (CBT) on nightmare frequency and to determine which kind of CBT is the most effective treatment. A systematic literature search was carried out in PsychInfo and PubMed articles published on or before May 1, 2008. The inclusion criteria were: nightmare treatment study, use of nonpharmacological treatment, not a qualitative case study, randomized-controlled trial (RCT). After selection, 12 peer-reviewed studies about 9 RCTs remained (2 follow-up studies and one displaying preliminary results). Several interventions have been reviewed including, recording one's nightmares, relaxation, exposure, and techniques of cognitive restructuring. The 12 evaluated articles varied in quality, and none fulfilled CONSORT guidelines. All articles used nightmare frequency as the primary dependent variable, and all found significant in-group differences (pre vs. post) for intervention or placebo (range d = 0.7-2.9). Five studies were able to find a significant group effect for the intervention compared to a waiting list control group. Only one study found significant differences between 2 intervention groups. Nightmare-focused CBT (exposure and imagery rehearsal therapy [IRT]) revealed better treatment outcomes than indirect CBT (relaxation, recording). IRT and exposure showed no meaningful differences, but only one RCT directly compared both techniques. Three different research groups demonstrated the effects of exposure, but only one group showed the effect of IRT. Thus, RCTs that compare IRT with exposure by independent research groups are much needed.