1 Twenty-four hours after ethanol withdrawal, dependent mice exhibited frequent head twitching. Naive mice exhibited similar twitching 15 min after treatment with 5-hydroxytryptophan (5-HTP) or 6 h after alpha-methyl-p-tyrosine (AMPT). Ethanol lessened the incidence of head twitches induced by any of these treatments. 5-HTP and AMPT each increased the incidence of head twitches induced by withdrawal of ethanol from dependent mice. 2 Drugs that affect the amount or activity of endogenous amines or cyclic nucleotides modified the incidence of head twitches. Nearly all drugs acted in the same direction on twitching elicited by any of these three treatments. 3 The incidence was lessened by: (a) methysergide, methergoline, MA 1420, p-chlorophenylalanine and p-chloroamphetamine; (b) dopamine, noradrenaline, L-DOPA, amphetamine and apomorphine; (c) hyoscine and nicotine; and (d) adenosine triphosphate, dibutyryl cyclic adenosine-3',5'-monophosphate (db cyclic AMP) and prostaglandins E1 and E2. 4 The incidence was increased by: (a) acetylcholine, carbachol and physostigmine; and (b) guanosine triphosphate, dibutyryl cyclic guanosine monophosphate (db cyclic GMP), theophylline and 3-isobutyl-1-methyl-xanthine. 5 These findings suggest that head twitching induced by these three treatments arises from a common biochemical mechanism, which may ultimately be a change in favour of cyclic GMP of the balance between this nucleotide and cyclic AMP within appropriate neurones. This imbalance appears to be elicited or increased by 5-hydroxytryptamine and acetylcholine and to be decreased by dopamine, noradrenaline and E prostaglandins. 6 Neither actinomycin D nor cycloheximide, given during the induction of ethanol dependence, altered the incidence of head twitches after ethanol withdrawal.