OBJECTIVE: To assess the 24-hour temporal-domain heart-rate variability correlates of treatment with fluoxetine or doxepin for depression. METHOD: A randomized evaluation of fluoxetine and doxepin measured a 50% change in the Hamilton Depression Rating Scale (HDRS) score as a response to therapy and was correlated with measures of standard deviation of the mean of all 5-minute segments of normal electrocardiographic R-R intervals (SDANN), standard deviation of all normal R-R intervals (SDNN), root mean square of successive differences in R-R intervals (r-MSSD), and percentage difference between adjacent normal R-R intervals that are greater than 50 msec (pNN50) from 24-hour electrocardiogram (ECG) tapes. RESULTS: Ten out of 14 patients responded. Response was associated with an increase in SDANN of 17% (P < 0.05). Nonresponse was associated with a 17% decrease in SDANN and a 22% decrease in SDNN (both P < 0.05). No other measures correlated with therapeutic response. No heart-rate variability (HRV) differences between the 2 drug therapies were observed. CONCLUSION: Twenty-four-hour HRV measures may be useful in assessing response to antidepressant therapy.