Following finding the acute and prophylactic efficacy of the anticonvulsants carbamazepine (CBZ) and valproate (VPA) we have documented loss of efficacy (presumably via tolerance) in 43% and 27% of patients, respectively. Algorithms including switching to drugs with different mechanisms of action and use of combination therapy are being explored. Positive effects of the anticonvulsant calcium channel blocker nimodipine have been observed and summarized in Z01 MH 02635. We have preliminarily identified clinical and biological markers of differential response among the mood stabilizers. These include rapid cycling as a predictor of poor response to both lithium and CBZ, but a good response to their combination. Temporal lobe hypermetabolism is associated with good response to CBZ, while frontal lobe hypometabolism appears to be a predictor of nimodipine response. In many instances response has been confirmed in individual patients in a placebo-drug-placebo-drug (B-A-B-A) design. This new methodology may prove more valuable than traditional designs, which are particularly ill suited to the variable presentations of bipolar illness, very expensive, and rarely funded. Patients who are inadequately responsive to one mood stabilizing anticonvulsant may respond to another. A blind, randomized, parallel group design protocol evaluating two new anticonvulsants with novel mechanisms of action against placebo - gabapentin and lamotrigine - has preliminarily shown good response on the CGI in six of the first 12 patients randomized to lamotrigine. The endogenous anticonvulsant neuropeptide TRH has been administered into the CSF during a routine lumbar puncture in eight patients. Attempts to convert TRH to a clinically relevant paradigm were initially achieved with daily parenteral TRH administration, although tolerance to the therapeutic effects developed in these treatment- refractory individuals over the course of several months. Significant improvement in mood, anxiety, and suicidal preoccupation was observed compared with the sham LP procedure. Repeated transcranial magnetic stimulation (rTMS) of the brain is being explored in randomized trials as a potential antidepressant modality. Following the first open study (in Neuroreport), significant improvement was observed during active but not sham phase in patients in a crossover trial. A randomized third study comparing 20 Hz, 1 Hz, and sham rTMS is in progress.