Depression is a common mood disorder. Although major ethical challenges make it nearly impossible to invasively and directly measure serotonin (5-hydroxytryptamine, 5-HT) levels in human brains, neuroimaging technologies have shown macroscopic structural and functional abnormalities in the prefrontal cortex (PFC) of depressed patients. The monoamine hypothesis of depression is based on the neurotransmitter imbalance, such as deceased serotonin brain levels are implicated in the cause of depression. Research has focused on the control mechanisms involved in the dorsal raphé nucleus (DRN) which is the serotonergic control center located in the midbrain. We hypothesized that activation 5-HT 2A receptor in PFC would increase serotonin levels by an AMPA-dependent mechanism in both DRN and PFC. Enhancement of the 5-HT in DRN may inhibit 5-HT level in PFC by 5-HT 1A receptor. This becomes the full feedback loop system. While 5-HT levels in the PFC have been well studied, pathway that modulate this DRN pool through upstream cascade interactions leading to a downstream feedback loop have been difficult to elucidate. Developing a mechanism-based pharmacokinetics (PK) and pharmacodynamics (PD) model to quantitatively describe the effect of 5-HT 2A receptors regulation to serotonin in the DRN and PFC would help us to better understand the complex brain. 5-HT 2A receptor agonist and AMPA receptor agonist and antagonist were used to activate or block the related receptor. Male Wistar rats underwent neurosurgery for implantation of microdialysis (MD) probes. Three to five rats were randomly assigned to experimental arms. Using the MD method, the drug combination was examined to explore the drug effect on time course of 5-HT release in DRN and PFC. Based on the experiment results, a mechanism-based PD model was developed. Phoenix WinNonlin ® and Berkeley Madonna™ were used for model estimation, external validation with secondary data set, and simulation. The result supports the possibility of a 5-HT 2A /AMPA feedback control circuit that originates in the PFC and modulates DRN and PFC 5-HT levels through feedback coupling of 5-HT. The time-course profiles of 5-HT in both DRN and PFC was well modeled and model parameters were estimated with good precision (CV% ranged from 1.37% to 35.03%). The mechanism model was developed to characterize and better understand the neurotransmitter mechanisms, providing estimations of various parameters of the disease related receptor system.