Objective To investigate the correlation between the occurrence of low fertilization in in vitro fertilization (IVF) cycles and sperm parameters. Methods This retrospective analysis was conducted among 264 infertile cases receiving of in vitro fertilization procedures in our Reproductive Medicine Center from June 2017 to October 2018. According to the presence or absence of two polar bodies after 5 to 6 hours of fertilization, they were divided into low fertilization group (fertilization rate < 30%) (n = 63) and normal fertilization group (fertilization rate ≥ 30%) (n = 201). The basic data and sperm parameters of the two groups were compared. The correlation was analyzed by binary logistic regression. The area under the curve (AUC) of Receiver Operating Characteristic Curves (ROC) was used to analyze the predictive value. Results Compared with the normal fertilization group, the head defected sperm rate (%H), sperm deformity index (SDI) and acrosome reaction rate (AR) in the low fertilization rate group were higher (P < 0.05). By binary logistic regression analysis, the correlation between three indicators and low fertility rate of IVF was determined: %H(Ol^R=1.132), SDI(Ol^R=24.826), AR(Ol^R=1.268) and established a regression equation that predicts the occurrence of low fertilization rate: LogitP = -19.234 + 0.124%H + 3.212SDI + 0.237AR (P < 0.05). Through ROC analysis, the regression equation with three parameters is compared with a single parameter, which has the greatest predictive value [AUC = 0.854, (95% CI: 0.805, 0.894)], the cut-off value is -1.461, the sensitivity was 85.71% and the specificity was 71.14%. Conclusion H %, SDI and AR were associated with the low fertilization rate of IVF. The predicted value of low fertilization rate (P) was calculated according to the logistic regression equation. When the predicted value of regression equation is >-1.461, patients were advised to choose ICSI to help pregnancy.