Abstract: AIM To investigate the association of structural features of patent foramen ovale (PFO) with occluder selection for transcatheter closure. METHODS Retrospective assessment of 98 consecutive patients who underwent PFO closure was made and the patients were divided into simple PFO group and complex PFO group according to the results of transthoracic echocardiography (TTE) and/or transesophageal echocardiography (TEE). Occluder size and efficacy were evaluated and compared between the two groups. RESULTS In the 98 patients (41 males, 57 females; aged from 8-74 years, mean age 41±14 years), the operation was successful in 92 cases, with the success rate of 93.9%. Of the 6 unsuccessful cases: 4 cases were failure of wire getting through the atrial septum, 1 case was pulmonary arteriovenous fistula and 1 case was pulmonary hypertension. Of the 98 cases, 42 cases were simple PFO and 56 cases were complex PFO. A 18/25 mm device was selected for most cases in simple PFO and 30/30 mm and 25/35mm devices were selected for most cases in complex PFO. Statistical significance between the two groups was assumed with a P value <0.01. Complete PFO closure as assessed by contrast TTE after the Valsalva maneuver at 6 months was achieved in 65(70.7%) of the patients, whereas a minimal, moderate, and large residual shunt persisted respectively in 20(21.7%) cases, 4(4.3%) cases and 3(3.3%) cases. Patients with 30/30mm and 25/35 mm devices had considerably large residual shunts. In complex PFO group, paroxysmal atrial fibrillation occurred in 1 case. CONCLUSIONT ranscatheter PFO closure is a safe and effective treatment, and the occluder selection for transcatheter closure is associated with the structural features of patent foramen ovale. A 18/25 mm device is selected for most simple PFO and 30/30 mm and 25/35 mm devices are used for most complex PFO.