We obtained Hamilton Rating Scale for Depression (HAM-D) scores and recorded 5 minutes of rhythm strip both before and after a therapeutic trial of antidepressant medications in 17 patients diagnosed with major depressive disorder (MDD). We calculated the standard deviation (SD) of interbeat intervals and the mean squared successive difference (MSSD) as measures of heart-rate variability (HRV). We then calculated Spearman rank-ordered correlation coefficients between the HRV measures and the HAM-D scores. Changes in SD and MSSD correlated with post-treatment HAM-D scores and with changes in HAM-D scores. These relationships were strongest in patients who responded positively to nontricyclic antidepressant medications. HRV before treatment was not predictive of treatment response, nor did HRV reliably reflect the severity of depressive symptoms. These findings indicate that pharmacologic treatment leading to improvement in MDD is associated with increased HRV. Hence, brief measures of HRV could be developed as a useful adjunctive, physiologic measure of treatment response to pharmacotherapy in clinical trials and other settings. Further, increased HRV associated with successful treatment of MDD may reflect improved autonomic function, decreasing the risk of cardiovascular mortality found in patients with MDD.