OBJECTIVE: It has been proposed that the inflammatory cytokine system is regulated through the vagus nerve, where vagal activation inhibits release of inflammatory cytokines and, therefore, inflammation. Thus, loss of vagal activation (i.e., reduced high-frequency heart rate variability [HF-HRV]) should result in greater inflammation. Evidence to date for this relationship has relied on animal models and resting states in humans. The present study used a psychosocial stressor to test whether stress-induced decreases in HF-HRV predict increases in circulating inflammatory markers. METHODS: Thirty healthy young women completed a speech stressor. HF-HRV was assessed before and during the stressor while circulating plasma interleukin 6, tumor necrosis factor alpha, and C-reactive protein were assessed before and 1 hour after the stressor. RESULTS: Consistent with the neural reflex for immunity, greater reductions in HF-HRV during the stressor were associated with greater increases in tumor necrosis factor alpha (beta = -0.29 to -0.47) and interleukin 6 (beta = -0.40 to -0.68) but not C-reactive protein (beta = 0.10 to 0.29) 1 hour after the stressor. CONCLUSIONS: These findings expand on the current literature by showing that changes in HF-HRV predict and precede changes in circulating inflammatory cytokines in humans and may have implications for treatment of inflammatory diseases.