To evaluate the effectiveness and safety of topiramate as add-on, long-term therapy for treatment-resistant bipolar-spectrum disorders, 34 DSM-IV bipolar-spectrum patients, including bipolar I (n = 28), bipolar II (n = 3), bipolar not otherwise specified (n = 2), and schizoaffective disorder bipolar type (n = 1), considered to be resistant to treatment with lithium, carbamazepine, and valproate, received increasing doses of topiramate as adjunctive therapy for their manic (n = 17), depressive (n = 11), hypomanic (n = 3), or mixed (n = 3) symptoms. Outcome measures included the Young Mania Rating Scale (YMRS), the Hamilton Rating Scale for Depression (HAM-D), and the Clinical Global Impression (CGI) for Severity. Patients were followed up for 6 months. Twenty-five patients (74%) completed the 6-month follow-up. Nine patients (26%) dropped out early due to lost of follow-up (n = 4), worsening of symptoms (n = 2), side effects (n = 1), hospitalization due to intercurrent illness (n = 1), and noncompliance (n = 1). By intent-to-treat analysis, there was a significant reduction in YMRS, HAM-D, and CGI scores (p < 0.0001 for all measures at the endpoint) after the introduction of topiramate. Most therapeutic effects appeared between weeks 2 and 6. Fifty-nine percent of manic patients and 55% of depressed patients were considered to be responders to the drug, which was well tolerated; only one patient discontinued due to side effects. The most common side effect was paraesthesia (n = 2). Ten patients experienced moderate weight loss during the follow-up period. The mean topiramate dose at endpoint was 202 +/- 65 mg/day. These preliminary results indicate that adjunctive topiramate may be useful in the long-term treatment of bipolar spectrum disorders, even in the most difficult-to-treat patients.