Objectives. The length of the tourniquet time during total knee arthroplasty (TKA) is related to the incidence of post-operative deep vein thrombosis (DVT). Our aim in this study was to investigate the effect of the early release of the tourniquet on the incidence of DVT in patients undergoing TKA. Methods. A total of 200 patients who underwent TKA between November 2015 and November 2016 were prospectively enrolled. The tourniquet was inflated before surgery and released immediately after the introduction of the components (early release group). This group was compared with a retrospective cohort of 200 primary TKAs, in which the tourniquet was released after the dressings had been applied (late release group). The presence of a DVT was detected using bilateral lower limb ultrasonography. Peri-operative clinical and follow-up data were collected for analysis. Results. The incidence of DVT in the early release group (9 of 196, 4.6%) was significantly lower compared with the late release group (24 of 200, 12%; odds ratio (OR) 0.35, 95% confidence interval (CI) 0.16 to 0.78, p = 0.008). The incidence of proximal DVT in the early release group (1 of 196 (0.5%)) was significantly lower than in the late release group (8 of 196, 4%; OR 0.12, 95% CI 0.02 to 0.99, p = 0.020). Although the mean intra-operative blood loss was higher in the early release group, the mean post-operative drainage, total blood loss, transfusion requirements and complications were not significantly different in the two groups. Conclusion. In patients who undergo TKA, releasing the tourniquet early is associated with a decreased incidence of DVT, without increasing the rate of complications. Cite this article: Bone Joint Res 2017;6:535–541