We read with interest and sympathy the recent paper of Aliyev and Aliyev (2008) reporting results of a double-blind placebocontrolled study in 40 patients with acute alcohol hallucinosis indicating valproate to be effective in reducing verbal hallucinations. This is the first placebo-controlled study in this rare disorder, a widely neglected subject of research. The pathophysiology of alcohol hallucinosis is not clear (Soyka, 1995;Soyka et al., 2000a). There is no evidence for a common genetic basis for alcohol psychosis and schizophrenia (Glass, 1989a). An impaired dopaminergic neurotransmission and hyperdopaminergic state has not been shown in alcohol hallucinosis (Soyka et al. 2000b). Recent PET findings indicate a hypofunction of the thalamus in patients with alcohol psychosis that may resolve on clinical improvement (Soyka et al., 2000(Soyka et al., , 2005Kitabayashi et al., 2007).With respect to pharmacological treatment, no standard therapy has been established yet . To date there are only a few predominant case series and reports, mostly uncited in the Aliyev paper, which may indicate a rather good prognosis in patients treated with neuroleptics (Soyka et al., 1992(Soyka et al., , 1997(Soyka et al., , 2007de Millas and Haasen, 2007). Therefore, it is surprising to read the results of this study. A few specific questions on this study raise some concern:1. Epidemiology: There is broad consensus that alcohol hallucinosis is a rare disorder (Tsuang et al., 1994). In the Psychiatric Hospital of the University of Munich (200 inpatients) over a 10-year episode we only saw 13 cases (Soyka et al., 2007). Recent data from official German hospital statistics indicate a prevalence of 0.6-0.7% in alcoholics (Soyka, 2008). We therefore wonder how a single centre is able to include 40 patients in an unreported period of time. This may either be explained by a much higher rate of alcohol hallucinosis in this area, as we understand a predominantly Islamic state compared to elsewhere, or schizophrenic patients with secondary alcoholism were included, a frequent diagnostic dilemma (Surawicz, 1980). Psychopathology is only briefly described in the Aliyev paper, but we wonder how these patients were excluded. 2. Methodology: There also is broad consensus that alcohol hallucinosis is a very acute disorder that deserves treatment. We wonder about the 2-week washout phasefrom what drugs? In our experience, most alcoholics with acute hallucinosis are admitted as emergency patients, free of medication and require acute treatment. In addition, none of the 40 patients recovered spontaneously within the 2 weeks although the prognosis is usually good (Glass, 1989b). This may again raise the question whether some patients with schizophrenia might have been included. 3. Outcome: Aliyev and Aliyev (2008) reported rating for verbal hallucinations of the PANSS scale as outcome. It is unclear from the paper how many patients were actually symptom free at the end of treatment. In the larger case series with neuroleptics, >90% were no l...