Risks of recurrence and bleeding are highest during the first weeks of anticoagulant therapy for venous thromboembolism (VTE). We therefore examined the early time course of recurrence and major bleeding in a pre-specified sub-analysis of the AMPLIFY trial, a randomised, double-blind, six-month comparison of oral apixaban with conventional therapy (enoxaparin followed by warfarin) in 5,395 patients with symptomatic proximal deep-vein thrombosis or pulmonary embolism. Early events were of particular interest because apixaban was given without initial heparin treatment. The primary efficacy and safety outcomes were the incidences of the adjudicated composite of recurrent symptomatic VTE or death related to VTE, and of adjudicated major bleeding, respectively. This analysis reports on recurrence and bleeding after 7, 21, and 90 days of therapy, in addition to the previously reported end-of-study results. These were the times specified before statistical analysis. Recurrent VTE after 7, 21, and 90 days, and six months had occurred in 18 (0.7%), 29 (1.1%), 46 (1.8%), and 59 patients (2.3%), respectively, given apixaban, and in 23 (0.9%), 35 (1.3%), 58 (2.2%), and 71 patients (2.7%), respectively, given conventional therapy. Major bleeding had occurred during these time intervals in 3 (0.1%), 5 (0.2%), 11 (0.4%), and 15 patients (0.6%), respectively, who received apixaban, and in 16 (0.6%), 26 (1.0%), 38 (1.4%), and 49 patients (1.8%), respectively, given conventional therapy. Efficacy of apixaban was non-inferior at each time point, with no excess of early recurrences. The reduced bleeding risk associated with apixaban began early during the course of treatment.