Percutaneous left atrial appendage closure (LAAC) has emerged as an alternative of stroke prevention in non-valvular atrial fibrillation (NVAF) patients. Peri-device leakage after LAAC is common. This retrospective, case–control study aimed to identify risk factors related with peri-device leakage after LAAC with Watchman devices. Patients who underwent Watchman devices implantation received trans-esophageal echocardiography (TEE) before, during and 45 days after procedure. Peri-device leakage was defined as a residual flow of any size detected with TEE. Patients with residual flows were compared with sex and age matched controls without leakage after implantation. Basic clinical characteristics, as well as LAA imaging characteristics were collected and compared. From 2014 to 2016, 125 consecutive patients were implanted with Watchman devices in our center. TEE at 45 days after implantation identified 53 patients with peri-device leakages (2.62 ± 1.55 mm), who were compared with 43 sex and age matched controls who also received the Watchman devices implantation and had no peri-device residual flow. The basic clinical characteristics, as well as LAA morphology were comparable between the two groups, while patients with leakages had larger LAA orifice, longer LAA body and larger LAA volume. Multivariate logistic regression analysis showed that LAA orifice size measured with TEE was the only independent risk factor predicting post-procedural leakage. The AUC of the receiver operating characteristic (ROC) curve was 0.70. Using the TEE orifice size cutoff value of 18.7 mm yielded a sensitivity of 0.92 (specificity 0.52), while the cutoff value of 23.1 mm yielded a high specificity of 0.91 (sensitivity 0.24). Minor peri-device leakage ( < 5 mm) is common after LAAC with Watchman devices. LAA orifice diameter measured with TEE is the independent risk factor predicting peri-device leakage after the implantation.