Lithium inhibits myo-inositol mono- and polyphosphatase activity in brain at concentrations similar to those optimal for the treatment of manic depressive psychosis. A consequence of this inhibition is the possibility that the availability of myo-inositol for the regeneration of polyphosphoinositides involved in cellular signalling mechanisms may be reduced. While there are no good models of manic depressive disorders in rodents, lithium is known to alter their behavioural responsiveness to a number of neurotransmitter receptor agonists, but the role of the phosphatidylinositol second messenger system in these effects is unknown. Consistent with the myo-inositol depletion hypothesis, when injected directly into the CNS, myo-inositol, but not its biologically inactive epimer, scyllo-inositol or D-mannitol, has been found to reverse a proconvulsant action of lithium in mice given the muscarinic receptor agonist, pilocarpine.