Adverse events reported in the context of medication administration may be due to pharmacodynamic and/or nonpharmacodynamic effects (eg, nocebo phenomena). Neurophysiological substrates of side effects may be examined in placebo-controlled antidepressant treatment trials. We explored the relationship between side effects and regional neurophysiologic changes in normal subjects receiving a 1-week placebo lead-in followed by 4 weeks randomized treatment with placebo (n ¼ 15) or venlafaxine IR (n ¼ 17). Quantitative electroencephalographic (QEEG) cordance measures were obtained before and during treatment, and side effects were assessed weekly using semistructured interviews. Side effect burden, characterized as the mean number of side effects per postrandomization visit, correlated significantly with neurophysiologic changes in the antidepressant group but not the placebo group. Medication group side effects were negatively correlated with changes in prefrontal cordance at end of placebo lead-in (r ¼ À0.67, po0.003), at 2 weeks (r ¼ À0.77, po0.002), and at 4 weeks (r ¼ À0.77, po0.004) post randomization. After controlling for the prefrontal change at the end of placebo lead-in, postrandomization brain changes did not further explain side effect burden. Changes in prefrontal brain function associated with later antidepressant side effects were observed during placebo lead-inFprior to the administration of medication. Prefrontal brain function during brief placebo administration may help explain susceptibility to the development of antidepressant side effects. Results of these exploratory hypothesis-generating analyses should be considered tentative until replicated.