OBJECTIVE To investigate the effectiveness of baclofen for the treatment of tics in children with Tourette syndrome (TS). BACKGROUND Baclofen, which contains both gamma-aminobutyric acid (GABA) and phenylethylamine moieties, was suggested in an open-label protocol to be an effective treatment for TS. This is a double-blind, placebo-controlled study to investigate this medication in children with TS. METHODS Subjects received, in a randomized sequence, 4-week medication cycles of baclofen (20 mg three times daily) and placebo with a 2-week intervening washout period between the cycles. Outcome measures included the Clinical Global Impression (CGI) scale, and the Yale Global Tic Severity Scale (YGTSS), the latter including subscales for total tics and overall impairment. Measures were assessed at baseline and on days 28, 42, and 70 of the study. RESULTS Ten children (seven boys and three girls, aged 8 to 14) with TS participated. Nine subjects completed the protocol; one dropped out for psychosocial reasons. No major side effects were reported. The mean change in CGI score (-0.9) after 4 weeks of baclofen treatment as compared with placebo treatment showed a significant improvement (95% CI, -1.7 to -0.1; p = 0.04). All subjects showed some amelioration in total YGTSS score during baclofen treatment. The mean change in total YGTSS score (-14.7) approached significance (95% CI, -30.3 to 0.9; p = 0.06). Examination of differences between baclofen and placebo treatment groups expressed as a percent change from baseline showed that baclofen had a statistically significant effect on both outcome measures. Subscales of the YGTSS showed that the reduction in total tic scores was primarily due to a reduction in the impairment score rather than a decrease in tics. CONCLUSIONS Children with TS may benefit from treatment with baclofen, although improvements may be related to factors other than tics. Larger studies directly comparing baclofen against other tic-suppressing agents are recommended.