We have much yet to accomplish in research on bipolar disorder. We must find vulnerability genes. We must identify the circuits that regulate mood, emotion, energy, and other relevant functions that are affected in bipolar disorder, and we must determine what goes wrong in those circuits during mania, depression, and other aspects of this illness. We will need to translate findings in basic neuroscience, genetics, and basic behavioral science into diverse clinical applications: novel treatments, diagnostic tools, epidemiologic approaches that could lead to preventive interventions, and surrogate markers for clinical trials. We must develop improved psychosocial interventions and test both pharmacologic and psychosocial treatments in trials that, simultaneously, improve the quality of care available and convince insurers and employers that these treatments are of substantial benefit and cost effective. The agenda is ambitious, but entirely feasible, given the scientific tools and technologies that are currently available or on the horizon. The National Institute of Mental Health is newly recommitted to harnessing these tools and technologies for the benefit of people with bipolar disorder.