The aims of this study were to establish optimal doses of doxycycline (dox) against Haemophilus parasuis on the basis of pharmacokinetic-pharmacodynamic (PK-PD) integration modeling. The infected model was established by intranasal inoculation of organism in pigs and confirmed by clinical signs, blood biochemistry, and microscopic examinations. The recommended dose (20 mg/kg b.w.) was administered in pigs through intramuscular routes for PK studies. The area under the concentration 0- to 24-hr curve (AUC ), elimination half-life (T ), and mean residence time (MRT) of dox in healthy and H. parasuis-infected pigs were 55.51 ± 5.72 versus 57.10 ± 4.89 μg·hr/ml, 8.28 ± 0.91 versus 9.80 ± 2.38 hr, and 8.43 ± 0.27 versus 8.79 ± 0.18 hr, respectively. The minimal inhibitory concentration (MIC) of dox against 40 H. parasuis isolates was conducted through broth microdilution method, the corresponding MIC and MIC were 0.25 and 1 μg/ml, respectively. The Ex vivo growth inhibition data suggested that dox exhibited a concentration-dependent killing mechanism. Based on the observed AUC /MIC values by modeling PK-PD data in H. parasuis-infected pigs, the doses predicted to obtain bacteriostatic, bactericidal, and elimination effects for H. parasuis over 24 hr were 5.25, 8.55, and 10.37 mg/kg for the 50% target attainment rate (TAR), and 7.26, 13.82, and 18.17 mg/kg for 90% TAR, respectively. This study provided a more optimized alternative for clinical use and demonstrated that the dosage 20 mg/kg of dox by intramuscular administration could have an effective bactericidal activity against H. parasuis.