An open-label pilot study explored the relationship between severity of depressive symptoms and venlafaxine dose required for clinical efficacy in outpatients with major depressive disorder (MDD). The utility of the neuroendocrine response to serotonergic (ipsapirone) and noradrenergic (clonidine) probes as predictors of venlafaxine dosage required for effective treatment was also explored. Nineteen medically healthy medication-free outpatients over 18 years of age who met criteria for MDD were studied. Participants received either a 20-mg dose of ipsapirone orally, a 0.002-mg/kg intravenous dose of clonidine, or placebo. Following a 1-week single-blind placebo lead-in, all subjects were treated with immediate release venlafaxine. Low-dose responders were defined as those subjects experiencing a >50% decrease in depression score on 37.5 mg, b.i.d., and high-dose responders were defined as those subjects experiencing similar improvement on venlafaxine doses of 75 mg, b.i.d., or higher. Subjects responding to low-dose treatment had a lower mean baseline Hamilton depression score than subjects requiring high-dose treatment. Neuroendocrine and temperature responses to clonidine or ipsapirone challenges were not significantly different in the high- vs. low-dose responders. Evaluation of models of "serotonergic-responsive" and norepinephrine-responsive" depression requires larger numbers of patients.