While we typically pick and choose among various drug classes to treat patients with borderline personality dis- order (BPD), we have a hard time bas- ing our choices on the evidence. Various reviews and meta-analyses of pharmacotherapy for BPD have reached diverging conclusions. A recent well-done meta-analysis reviewed three classes of medications—antipsychotics, mood stabilizers, and antidepressants— in the treatment of BPD and schizo- typal personality disorder. The main antipsychotics surveyed included haloperidal, olanzapine, risperidone, and aripiprazole; the main mood stabi- lizers were carbamazepine, valproate, and topiramate; and the main antide- pressants were fluoxetine, amitriptyline, and phenelzine. A total of 21 trials were included, most of which focused on BPD and had small sample sizes of between 20 and 60. On average, antipsychotics improved cognitive-per- ceptual symptoms (eg, suspiciousness, depersonalization), anger, and global functioning moderately more than placebo, but had no significant effect on impulse control, depressed mood, or anxiety. Mood stabilizers had large effects on impulse control, anxiety, and anger, and moderate effects on depres- sion and global functioning. Antidepressants yielded significant but small effects only on anxiety and anger (and not depression). (Ingenhoven T et al., J Clin Psychiatry online ahead of print.)