The purpose of this study is to determine the safety and effectiveness of a long-acting injectable formulation of risperidone in stable bipolar patients randomly switched from their current add-on oral antipsychotic (olanzapine, risperidone, or quetiapine) therapy to long-acting injectable risperidone. The patients switched to long-acting injectable risperidone will be compared to patients who continue on their oral antipsychotic treatment regimen

This an open-label, randomized study. Approximately 40 stable bipolar patients who are on an atypical antipsychotic (olanzapine, risperidone, quetiapine) plus adjunct bipolar treatment consisting of (a maximum or two of lithium, valproate or lamotrigine, and, if applicable, one antidepressant) will be randomized to two arms. In one arm, 25 milligrams of long-acting injectable risperidone will replace the oral atypical antipsychotic as adjunct therapy and in the other arm, patients will continue with their current atypical antipsychotic therapy. Trial duration is 6 months. In the long-acting injectable risperidone arm, the oral atypical antipsychotic will be continued (as supplementation) for 3 weeks after the first injection of long-acting risperidone and then discontinued. Investigators, based upon the patient's response, may increase the dose of injectable risperidone to 37.5 mg after 6 weeks on the 25-mg dose and to 50 mg after at least 4 weeks on the 37.5-mg dose. Risperidone oral supplementation is allowed. In the oral antipsychotic only arm, the oral atypical antipsychotic dose can also be increased as required. The primary efficacy outcome will be measured by changes in the Clinical Global Impression - Severity of Illness subscale (CGI-S), from baseline to endpoint, and will be compared between the treatment groups. Safety will be monitored throughout the study. The primary hypothesis is that patients switched to long-acting injectable risperidone will be able to tolerate this formulation of risperidone and maintain or even improve their reduction in bipolar symptomatology compared with baseline, and compared with subjects who continue in the oral antipsychotic arm. The secondary hypothesis is that patients switched to long-acting injectable risperidone will have a longer time to intervention for a mood episode (either mania or depression) as compared with subjects who continue in the oral antipsychotic arm. Risperidone, formulated for intramuscular injection, 25 mg every 2 weeks. Patients treated with injectable risperidone continue their original oral atypical antipsychotic (AAP) dose for 3 weeks. Investigators, at their discretion, may increase the dose of injectable risperidone to 37.5 mg after 6 weeks on the 25-mg dose and to 50 mg after at least 4 weeks on the 37.5-mg dose. Study duration is 6 months.

Inclusion Criteria:

• Stable outpatients meeting the DSM-IV criteria for Bipolar I or Bipolar II Disorder
• YMRS score of <= 19, MADRS score <= 19 and the Clinical Global Impression - Severity of Illness subscale (CGI-S) score <= 4 at screening and baseline
• Must be receiving stable doses of one oral atypical antipsychotic (olanzapine, risperidone, or quetiapine) in combination with a maximum of two of lithium, valproate or lamotrigine, and, if applicable, one antidepressant)
• Subject is healthy on the basis of a pre-trial physical examination, medical history and the results of blood biochemistry, hematology tests or urinalysis tests within 2 weeks of randomization (i.e. during screening)
• Female subjects must be postmenopausal (for at least 1 year), surgically sterile, or practicing an effective method of birth control before entry and throughout the study, and have a negative urine pregnancy test at screening and baseline

Exclusion Criteria:

• Have a serious unstable medical illness
• Had previous treatment with a long-acting injectable antipsychotic medication
• Known to be a risperidone non-responder or have a confirmed or suspected history of hypersensitivity or allergy to risperidone
• Patients at imminent risk of injury to self or others, or of causing significant damage to property
• Current drug or alcohol dependence