Lithium and valproate are commonly used mood stabilizers, but their action pathways are not clearly understood. They also suffer from multiple toxic effects that limit their utility. Elucidating their action mechanisms could lead to newer agents and better understanding of the etiopathogenesis of bipolar disorder. We have expanded the study of signaling mechanisms of lithium and valproate by using Drosophila circadian locomotor activity as a robust behavioral assay that is amenable to genetic manipulations. We demonstrate that lithium affects the circadian system of Drosophila similarly to what has been reported in the mammalian studies. We show that lithium and valproate share effects on the circadian locomotor activity of Drosophila: they lengthen the period of circadian rhythms and increase arrhythmicity. Valproate exerts these effects in a weaker fashion than does lithium. We also tested the circadian alterations in multiple mutant lines of Drosophila bearing defects in the GSK-3beta gene and other clock genes in response to lithium administration. We show that lithium partially rescues the shortening of circadian period when the GSK-3beta gene is overexpressed only in specific circadian pacemaker neurons, thus implicating GSK-3beta as a component in lithium's effect on the circadian oscillator. Moreover, lithium also lengthens the period in GSK-3beta heterozygous mutants and doubletime long mutants. These results establish a basis for using Drosophila genetics to investigate more fully lithium and valproate action mechanisms.