Objective To investigate the role of the perfusion index (PI) in predicting body movement during indolent induced abortion with intravenous anesthesia by multivariate logistic analysis. Methods 68 early pregnancy women, ASA I, aged 20 y - 30 y, weighed 45 kg - 60 kg, were subjected to indolent induced abortion with 1.8 mg/kg propofol and 0.5 mg/kg ketamine in this study and allocated into moving group and non-moving group based on whether body movement responses occurred during the surgical procedures. The PI, bispectral index (BIS), HR, SpO2 and blood pressure were monitored before induction of anesthesia (T1) and 3 min after induction (T2). Results 32.8% of the pregnant women took place body movement during the procedures. There were no significant difference between two groups in age, weight, gestational weeks, the history of induced abortion or labor, BIS index, blood pressure and HR at T1 and T2 (P>0.05). Compared with non-moving group, the pregnant women in moving group showed lower PI at T1 and T2 (0.39±0.16, 0.99±0.40) and the ratio of PI at T2/T1 (PI(T2/T1)) (2.91±1.41) (P<0.05). Logistic analysis showed that PI at T2 and PI(T2/T1) were independent factors for predicting body movement during the induced abortion. According to the analysis of receivers operating characteristic curve, using PI at T2 and PI(T2/T1) for predicting body movement, the areas under curves (AUC) were respectively 0.897 and 0.857. Conclusions Measurement of PI is helpful for monitoring anesthesia, and PI at T2 as well PI(T2/T1) might be effective in predicting body movement during the induced abortion.