Pharmacokinetics (PK)-pharmacodynamics (PD) modeling, the mathematical description of the relationship between PK and PD, can estimate and predict relevant parameters associated with onset, magnitude and time courses of dose-concentration-effect of a drug. In this report, we introduce a new nonsteady-state and time-dependent PK-PD modeling of a single dose of morphine in which time courses of concentration of unconjugated and estimated conjugated morphine in compartments of either plasma or biophase (cerebrospinal fluid, CSF) and multiple anti-nociceptive effects across thermal and mechanical stimulus modalities in rats were studied. The results showed that: (1) both intragastric and intraperitoneal administration of a single dose of morphine resulted in a differential anti-nociceptive effect in both magnitude and time course of the drug between thermal and mechanical painful stimuli (anti-mechanical pain effect was 2-3 fold stronger than anti-thermal pain effect, P < 0.01); (2) the PK data showed that the area under concentration-time curves of conjugated morphine was 4.5 and 2.0 fold bigger than unconjugated morphine in either plasma and biophase compartments, suggesting that the PK processes of unconjugated morphine are different from that of conjugated morphine; (3) the PD data also showed a change in PD characteristics of unconjugated and conjugated morphine across systemic and biophasic compartments for anti-mechanical pain effect, while there was no change at all for anti-thermal pain effect; (4) the difference in analgesia of a single dose of morphine across thermal and mechanical stimulus modalities was well reflected by the difference in the nonsteady-state and time-dependent PK-PD modeling, namely, the clockwise hysteresis loop model well represents the relationship of the time course between unconjugated/conjugated morphine concentration (both plasma and biophase) and anti-thermal pain effect, while the counter-clockwise hysteresis loop model well represents that between conjugated morphine concentration (mainly in biophase) and anti-mechanical pain effect. Taken together, the multiple PD-PK modeling is more useful in estimation and prediction of onset, magnitude and time courses of concentration-multiple pharmacological effects of morphine than simple PK or PD models, and establishment of various multiple PD-PK modeling might also be more useful in optimizing clinical use of existing drugs as well as new drugs for analgesia or treatment of other diseases.