Post-traumatic stress disorder (PTSD) is a highly prevalent (7.8% lifetime rate) anxiety disorder with impairment in daily functioning, frequent suicidal behaviour and high rates of co-morbidity. Fortunately, PTSD is responsive to pharmacotherapy and psychotherapy. The selective serotonin reuptake inhibitors (SSRIs) are the most studied medications for PTSD, with the largest number of double-blind, placebo-controlled trials. Of the SSRIs, sertraline, paroxetine and fluoxetine have been the most extensively studied, with sertraline and paroxetine being US FDA-approved for PTSD. These studies have demonstrated that SSRIs are effective in short-term trials (6-12 weeks). Furthermore, continuation and maintenance treatment for 6-12 months decrease relapse rates. Besides being the most studied and effective drugs for PTSD, SSRIs have a favourable adverse effect profile, making them the first-line treatment for PTSD. If SSRIs are not tolerated or are ineffective, non-SSRIs should be considered. Serotonin-potentiating non-SSRIs, such as venlafaxine, nefazodone, trazodone and mirtazapine, have been evaluated in PTSD only in open-label and case studies. Because of their promising results and relatively good safety profile, they should be considered as second-line treatment. Monoamine oxidase inhibitors (MAOIs) and tricyclic antidepressants (TCAs) have both been evaluated in a small number of double-blind, placebo-controlled studies. The results have been inconsistent but promising. In the limited comparative studies, MAOIs appeared superior to TCAs but patients continued to have residual symptoms. These drugs have significant adverse effects, such as cardiovascular complications, and safety issues, such as ease of overdose. Therefore, TCAs and MAOIs should be considered as third-line treatment. Anticonvulsants have been evaluated in PTSD in open-label studies and results have been positive for carbamazepine, valproic acid, topiramate and gabapentin. A small double-blind, placebo-controlled study demonstrated efficacy of lamotrigine for PTSD. Anticonvulsants should be considered where co-morbidity of bipolar disorder exists, and where impulsivity and anger predominate. Bupropion (amfebutamone), a predominantly noradrenergic reuptake inhibitor, was ineffective in PTSD in an open-label study. Benzodiazepines were ineffective in a double-blind, placebo-controlled study despite encouraging case reports. They should be avoided or used only short term because of potential depressogenic effects, and the possibility that they may promote or worsen PTSD. Buspirone, a non-benzodiazepine anxiolytic, was found to be effective in PTSD only in open-label studies. Recently, atypical antipsychotics were as effective as monotherapy and as an augmenter to SSRIs in open-label/case studies and small double-blind, placebo-controlled trials; atypical antipsychotics should be considered in PTSD where paranoia or flashbacks are prominent and in potentiating SSRIs in refractory cases.