Lithium is considered to be the first choice mood stabilizer in recurrent mood disorders. Its widespread and large-scale use is the result of its proven efficacy. In spite of this fact, patients have been observed to show a variable response to lithium treatment: in some cases it is completely effective in preventing manic or depressive relapses, while in other cases it appears to show no influence on the disease course. The possible definition of a genetic liability profile for adverse effects and efficacy will be of great help, as lithium therapy needs at least 6 months to be effective in stabilizing mood disorders. During the last few years, a number of groups have reported possible liability genes. Lithium long-term prophylactic efficacy has been associated with serotonin transporter protein, tryptophan hydroxylase and inositol polyphosphate 1-phosphatase variants. A number of other candidate genes and anonymous markers did not yield positive associations. Therefore, even if some positive results have been reported, no unequivocal susceptibility gene for lithium efficacy has been identified. Although the available data may not currently allow a meaningful prediction of lithium response, future research is aimed at the development of individualized treament of mood disorders, including the possibility of ‘pharmacological genetic counseling’.