The purpose of the present study is to understand the pharmacokinetic feature of mefloquine measured by erythrocytes and plasma in Schistosoma japonicum (S. j.)-infected mice and non-infected mice after oral administration of the drug at single doses. A high-performance liquid chromatography (HPLC) method was used to measure the plasma and erythrocyte concentrations of mefloquine at varying intervals posttreatment. Our results demonstrated that in non-infected mice treated orally with mefloquine at an ineffective dose of 50 mg/kg or effective dose of 200 mg/kg for 2-72 h, the erythrocyte-to-plasma ratios of mefloquine were 5.8-11.2 or 2-14.2. On the other hand, in S. j.-infected mice treated with the same single doses of the drug, the erythrocyte and plasma drug concentration ratios were 3.1-4.6 or 2.9-8.5, manifesting that either in infected mice or in non-infected mice that received oral mefloquine resulted in higher concentration of mefloquine in erythrocytes than that in plasma. Unexpectedly, under oral administration of mefloquine at a higher single dose of 200 mg/kg, the pharmacokinetic parameter C max values for plasma from S. j.-infected and non-infected mice were 1.6 ± 0.3 and 2.0 ± 0.4 μg/mL, respectively, which were below the determined in vitro LC50 (50 % lethal concentration) value of 4.93 μg/mL. Therefore, the plasma concentration of mefloquine may display a little effect against schistosomes during the treatment. Although the values of T 1/2 and AUC0-∞ for erythrocytes were significantly longer and higher in infected mice than those of corresponding non-infect mice that received the same single mefloqine dose of 50 mg/kg, the C max value was only 2.6 ± 0.4 μg/mL lower than the determined in vitro LC50, which may explain why this low single dose is ineffective against schistosomes in vivo. After administration of higher mefloquine dose of 200 mg/kg, the C max value for erythrocytes in infected mice was 30 % (7.4 ± 0.7 versus 10.7 ± 2.7 μg/mL) lower than that in the corresponding non-infected mice, but its level was above the determined in vitro LC95 (95 % lethal concentration) value of 6.12 μg/mL. Meanwhile, longer T 1/2 value of 159.2 ± 129.3 h in infected mice led to significant increase in AUC0-∞ value (1969.3 ± 1057.7 vs 486.4 ± 53.0 μg/mL·h), relative to corresponding non-infected mice. In addition, the mean residence time (MRT0-∞) in infected mice was also significantly longer than that in non-infected mice. All these results may beneficial for the treatment. According to the results, we suggest that higher ratios of mefloquine concentration in erythrocytes to plasma may offer a way to transport mefloquine to the worm gut through ingestion of erythrocytes by the worms, where the gut is the site for displaying the effect by mefloquine.