Introduction: Current guidelines report no benefit for patent foramen ovale (PFO) closure compared to medical treatment in patients with cryptogenic ischemic stroke (IS) or transient ischemic attack (TIA). Two recent randomized-controlled clinical trials (RCTs) have challenged these recommendations. We performed a systematic review and network meta-analysis (NMA) of RCTs to estimate the safety and efficacy of closure compared to medical treatment, and to compare available devices. Methods: We conducted pairwise meta-analyses for closure vs medical therapy, and for each device vs medical therapy. We performed subgroup analyses for the outcome of IS recurrence according to shunt size and atrial septal aneurysm (ASA) presence. For all devices we calculated the pooled rates of effective closure and serious peri-procedural complications. We used NMA to combine direct and indirect evidence for any given pair of treatments. Results: Literature search identified 6 eligible studies. PFO occlusion was associated with reduced risk of recurrent IS (RR=0.39, 95%CI:0.18-0.88) and IS/TIA (RR=0.66, 95%CI:0.46-0.94), but with increased risk of new-onset atrial fibrillation (AF; RR=4.73, 95%CI:2.09-10.71) compared to medical treatment. We detected no differences on the risk of recurrent IS according to PFO size (p=0.83) or ASA presence (p=0.39). Amplatzer and GORE HELEX/CARDIOFORM occluders were associated with more effective IS/TIA prevention (Panels A-B) and lower AF risk compared to STARFlex Septal Occluder (Panel C). Conclusions: PFO closure is related with reduced risk of recurrent IS/TIA and with increased risk of new-onset AF. These associations are independent of PFO size and ASA presence.