In a group of 17 patients with non-SAD depressive disorders we compared the response of bipolar spectrum versus unipolar patients to treatment with light therapy. The main hypothesis was that bipolar spectrum depressed patients would preferentially respond to bright light therapy as compared to unipolar depressed patients. All patients were treated with either 400 or 2500 lux phototherapy for 2 h on seven consecutive days. All outcome measures, which included the SIGH-SAD, CGI, and the Anxiety and Depressive Factors of the SCL-90, showed significant improvement in the bipolar vs. the unipolar spectrum patients. Unexpected this occurred regardless of the intensity of the light. These changes were judged to be quite clinically significant. All patients showing response were noted to have maintained their response at a 3-month follow-up.