Valproate is used in psychiatry as a mood-stabilizer and has efficacy in controlling impulsivity, therefore it seems to be a useful pharmacologic agent in pathologic gambling.

To assess the efficacy of valproate in the treatment of pathologic gambling.

A group of 13 patients, 9 male and 4 female, mean age 32.6, diagnosed with pathologic gambling, according to the DSM-IV-TR criteria, received treatment with flexible dose of valproate (600–1400 mg/day, mean dose 850 mg/day) and 2 times/week cognitive therapy sessions. Patients with other axis I and II conditions were excluded. Also, none of these patients received valproate or structured cognitive therapy prior to this trial. Patients were monitored for 3 months using monthly administered Gambling Symptoms Assessment Scale (G-SAS), Clinical Global Impressions (CGI) and 10 points self-evaluated Visual Analogue Scale (VAS).

Valproate was associated with a mean decrease of 10.5 points on G-SAS at week 12, compared to baseline (p < 0.05), while the CGI decreased with only 1.3 points (p = 0.122) and the VAS improved with 2.3 points (p < 0.05). The combined treatment was associated with good tolerability profile, since no drop-out was recorded in the study group due to side events and only 7 reports of mild and medium adverse events were recorded.

Valproate treatment associated with cognitive therapy could be a useful strategy in the management of pathologic gambling. Valproate's tolerability profile is good and the cognitive therapy offer an opportunity to ventilate the negative automatic thoughts, emotional distress and low impulse control strategies.