Objective To evaluate the in vivo efficacy against Babesia microti of two therapies commonly used in the clinic: a combination of atovaquone (ATQ) with azithromycin (AZM) and malarone in immune-normal BALB/c and non-obese diabetic/severe combined immunodeficient(NOD/SCID) mice models. Methods In total, 69 BALB/c and 15 NOD/SCID mice were each inoculated with 10⁷ Babesia microti-infected erythrocytes. The mice of each immune type were divided into three groups: ATQ + AZM group (195 mg/kg ATQ + 32.5 mg/kg AZM), malarone group (62.5 mg/kg ATQ + 25 mg/kg proguanil), and control group (5% soluble starch solution), the drug was administered by gavage of 0.2 ml/10 g body weight. In the drug-suppression testing in BALB/c mice(12 mice each of two drug groups, and 15 mice in the control group), the mice received the first dose of drug 4 h after infection, and then the dosing was continued for 10 consecutive days. Three mice were randomly selected for blood sampling from tail tip before gavage administration on 1, 3, 5, 7, and 10 days post-infection. Thin blood smears were prepared for microscopy to examine B. microti infection of erythrocytes, and estimate the erythrocyte infection rate (EIR); the blood samples were tested for 18S rRNA using qPCR to examine the gene value. In the drug-therapy testing experiment with BALB/c mice (10 mice each of 3 group), tail tip blood samples were collected from all mice 7 d post-infection to determine whether the infection was established, then when the infection was confirmed, the drug administration was initiated and continued for 10 consecutive days; microscopic examination was conducted for the blood samples collected on days 10, 11, 12, 13, 15, 17, 19, 24 and 27 post-infection to estimate the EIR. On 27 d post-infection, 5 mice randomly selected from each group were intraperitoneally injected with immunosuppressant dexamethasone sodium phosphate solution (200 μl/mouse) for 7 consecutive days; starting on 3 d after immunosuppressant injection, daily blood samples were examined by staining microscopy to observe recrudescence; Orbital sinus blood samples were collected from the mice randomly selected (5 from each group), and the anticoagulated whole blood samples obtained were pooled and injected intraperitoneally to corresponding groups of normal BALB/c (5 mice/group); on 7-10 d post subinnoculation, blood samples were examined to estimate EIR. In the experiment with NOD/SCID mice (3 groups, 5 mice each group), drugs were administered from 10 d post infection and continued for 10 days. Tail tip blood samples of the mice of three groups were collected on 10, 12, 15, 17, 19, 21, 24, 27, 29, 31, 35, 42 and 49 d post-infection, and examined by staining microscopy to estimate EIR. Results In the drug-suppression test in BALB/c mice, both ATQ + ATM and malarone obviously suppressed parasitemia. Microscopic examination indicated the only one mouse in the ATQ + AZM group was found infected on 3 d and 5 d post infection, with EIR of (0.20 ± 0.12)% and (0.30 ± 0.17)% respectively, while decreased to 0 on 7 d post-infection (8 d post-dosing). EIR was 0 in all mice in the malarone group. Significant differences were observed in EIR of control vs ATQ + ATM group and control vs malarone group (F = 151.6, 153.5, P < 0.05). The relative value of 18S rRNA gene was at 0.010 2 ± 0.001 2 and 0.007 8 ± 0.006 6 in malarone and the ATQ + AZM groups, respectively on 7 d post infection, which was significantly different from the control group (68.143 8 ± 79.122 9) (F = 7.376, 7.382, P < 0.05). On 10 d post-infection (1 d upon drug withdrawal), the value of 18S rRNA gene decreased to 0.001 7 ± 0.000 9 and 0.000 8 ± 0.000 6 in the malarone and ATQ + AZM groups on day 10 post-infection, respectively, which were significantly different from the control group (t = 4.229, 4.229, P < 0.05). In the drug-therapy tests with BALB/c, EIR in control, ATQ + AZM, and malarone groups peaked on 7 d post-infection, (36.67 ± 10.85)%, (35.30 ± 6.46)%, and (33.53 ± 7.37)% respectively, and decreased to(10.47 ± 8.02)%, (1.53 ± 0.31)% and (6.27 ± 1.01)%, respectively, on 11 d post-infection (5 d upon drug withdrawal); EIR in all groups declined to 0 on 15 d post-infection. B. microti was observed in one mouse in the control group 3 d after immunosuppressant injection, and seen in both the ATQ + AZM and malarone groups 5 d after immunosuppressing, indicating that recrudescence occurred in all groups. Subinoculation experiment showed that three mice that received blood from either the ATQ + AZM or malarone groups developed parasitemia 7 d after subinoculation. One mouse and two mice showed parasitemia in the ATQ + AZM and malarone groups, respectively, 9 d after subinoculation, and no parasite was observed in any mice from 10 d post subinoculation. In the test with NOD/SCID mice, the EIR of ATQ + AZM and malarone groups rapidly reduced after treatment from its peak value on 10 d post infection [(59.90 ± 0.10) % and (59.37 ± 0.35)%, respectively] to 0 on 24 d post-infection, whereas B. microti was observed again on day 29 and 42 post infection. EIR in the control group fluctuated between (47.20 ± 0.80)% and (66.80 ± 0.80)%, and all mice died on 45 d post-infection, of which the process was significantly different from ATQ + ATM, and malarone group (F = 5 505 and 5 984, respectively, P < 0.05). Conclusion Both ATQ + ATM and malarone that commonly used in clinic showed suppressive effect to some extent on the proliferation of B. microti in infected mice, but could not clear up all the parasites; the blood of mice remains infective after drug treatment and recrudescence may occur in immunocompromised hosts, causing high infection rate of erythrocyte.