OBJECTIVE： To establish a method for the determination of levofloxacin concentration in human pleural effusion， to study its pharmaceutical characteristics. METHODS： Totally 6 patients with infectious pleural effusion received levofloxacin 0.4 g qd intravenous drip. Pleural effusion was collected at 0.5， 1， 2， 4， 8， 12 and 24 hours after administration. After treated with methanol precipitation protein， HPLC was used to determine the concentration of levofloxacin. The determination was performed on Agilent ZORBAX SB-C18 column with mobile phase consisted of methanol-0.02 mmol/L KH2PO4 buffer （containing 0.3％ triethylamine， 70 ∶ 30， V/V at the flow rate of 1.0 mL/min. The detection wavelength was set at 294 nm， and the column temperature was 35 ℃. The sample size was 20 μL. The pharmacokinetic parameters were calculated with WinNonlin 5.2 software. RESULTS： Under this chromatogram condition， retention time of levofloxacin was about 4.9 min， the peak shape was good， the baseline was stable， and the determination of endogenous substances in pleural effusion had no interference. The linear range of levofloxacin were 0.625-20 μg/mL（R2＝0.998 9）. The relative recovery rates were （83.75±1.66）％-（87.73±2.43）％ for low， medium and high concentration samples （n＝3）； RSDs of intra-day were 2.23％-4.96％ （n＝5）； RSDs of inter-day were 4.10％-4.78％（n＝5）； the accuracy ranged （97.76±4.85）％-（100.87±2.25）％（n＝5）； RSD of concentration was no more than 5％ in stability test （n＝3） for low， medium and high quality control sample. The pharmacokinetic parameters of levofloxacin included cmax were （2.21±0.87） μg/mL； AUC0-24 h were （37.31±11.94） μg·h/mL； t1/2 were （4.50±0.21） h. CONCLUSIONS： Established method is simple， reliable and sensitive， and can be used for the determination of levofloxacin concentration in human pleural effusion and its pharmacokinetic study.