Objective To investigate the effect and safety of time of temporarily-closed wound drainage on blood loss of primary total knee arthroplasty (TKA) after intravenous and intra-articular injection of tranexamic acid (TXA). Methods Eighty female patients were selected from 102 patients who underwent primary TKA between September 2015 and July 2016, who were randomly divided into 4 groups: control group (group A), 30 minutes group (group B), 60 minutes group (group C), and 90 minutes group (group D), 20 patients each group. No significant difference was found in age, body mass index, side, pathogen, duration, and preoperative hemoglobin, albumin, and hematocrit between 4 groups (P>0.05). All the patients received intravenous injection of 1 g TXA at 10 minutes before removing the tourniquet. The patients in group A were injected with 60 mL normal saline into the articular cavity and closed drainage after surgery, while the patients in groups B, C, and D were injected with 60 mL TXA into the articular cavity and closed drainage for 30, 60, and 90 minutes respectively. The volume of drainage at 24 hours after operation, the total blood loss, the postoperative hemoglobin level, maximum hemoglobin loss, albumin loss, the volume and frequency of blood transfusion, venous thrombo embolism rate, and pulmonary embolism rate were recorded and compared between groups. Results The volume of drainage and total blood loss in groups B, C, and D were less than those of group A, showing significant difference between groups C, D and group A (P<0.05), but no significant difference between group B and group A (P>0.05). The volume of drainage at 24 hours after operation in group B was higher than that in groups C and D, showing significant difference between groups B and D (P<0.05), but no significant difference was found between groups C and D (P>0.05). There was no significant difference in the total blood loss between groups B, C, and D (P>0.05). The hemoglobin loss and albumin loss gradually decreased from groups A to D, but no significant difference was found between groups (P>0.05). No venous thrombo embolism and pulmonary embolism occurred. The hemoglobin value decreased to 28 g/L at 3 days after operation in 1 patient of group D, who received venous transfusion of 20 g human albumin. Conclusion Intravenous and topical application of TXA in TKA can significantly decrease postoperative bleeding. Topical TXA combined with 60 minutes temporarily-closed wound drainage may reduce postoperative blood loss to the greatest extent without increasing the risk of venous thrombo and pulmonary embolism event after TKA．