A randomized controlled trial shows that lithium treatment in bipolar disorder patients reduces daytime activity, advances its onset time, and increases volatility in circadian rest-activity patterns. These changes appear early and independently of mood, offering promise as early biomarkers of treatment response.