Objectives: To test the effectiveness of a television-supported self-help intervention for problem drinking. Methods: Dutch television viewers (N = 181) drinking in excess of the guidelines for low-risk alcohol use were randomly assigned either to the Drinking Less TV self-help course (consisting of five televised sessions supplemented by a self-help manual and a self-help website) or to a waitlisted control group. To ensure trial integrity, intervention delivery was mimicked beforehand by sending intervention participants weekly DVDs in advance of the actual telecasts in 2006. Pre-post assessments were carried out on both groups, as well as a 3-month follow-up assessment on the intervention group to study effect maintenance. The primary outcome measure was low-risk drinking. Results: The intervention group was more successful than the waitlist group in achieving low-risk drinking at post-intervention (OR = 9.4); the effects were maintained in the intervention group at 3-month follow-up. Conclusions: The low-threshold television-based course Drinking Less appears effective in reducing problem drinking. © 2009 Elsevier Ltd. All rights reserved.