BackgroundAlthough repetitive transcranial magnetic stimulation (rTMS) has been proven to be effective in the upper limb motor function and activities of daily living (ADL), the therapeutic effects of different stimulation protocols have not been effectively compared. To fill this gap, this study carried out the comparison of the upper limb motor function and ADL performance of patients with stroke through a network meta-analysis.MethodsRandomized controlled trials (RCTs) on the rTMS therapy for stroke were searched from various databases, including PubMed, web of science, Embase, Cochrane Library, ProQuest, Wanfang database, the China National Knowledge Infrastructure (CNKI), and VIP information (www.cqvip.com). The retrieval period was from the establishment of the database to January 2021. Meanwhile, five independent researchers were responsible for the study selection, data extraction, and quality evaluation. The outcome measures included Upper Extremity Fugl-Meyer Assessment (UE-FMA), Wolf Motor Function Test (WMFT), Modified Barthel Index (MBI), the National Institute of Health stroke scale (NIHSS), and adverse reactions. The Gemtc 0.14.3 software based on the Bayesian model framework was used for network meta-analysis, and funnel plots and network diagram plots were conducted using Stata14.0 software.ResultsNinety-five studies and 5,016 patients were included ultimately. The intervention measures included were as follows: placebo, intermittent theta-burst stimulation (ITBS), continuous theta-burst stimulation (CTBS),1 Hz rTMS,3–5 Hz rTMS, and ≥10 Hz rTMS. The results of the network meta-analysis show that different rTMS protocols were superior to placebo in terms of UE-FMA, NIHSS, and MBI outcomes. In the probability ranking results, ≥10 Hz rTMS ranked first in UE-FMA, WMFT, and MBI. For the NIHSS outcome, the ITBS ranked first and 1 Hz rTMS ranked the second. The subgroup analyses of UE-FMA showed that ≥10 Hz rTMS was the best stimulation protocol for mild stroke, severe stroke, and the convalescent phase, as well as ITBS was for acute and subacute phases. In addition, it was reported in 13 included studies that only a few patients suffered from adverse reactions, such as headache, nausea, and emesis.ConclusionOverall, ≥10 Hz rTMS may be the best stimulation protocol for improving the upper limb motor function and ADL performance in patients with stroke. Considering the impact of stroke severity and phase on the upper limb motor function, ≥10 Hz rTMS may be the preferred stimulation protocol for mild stroke, severe stroke, and for the convalescent phase, and ITBS for acute and subacute phases.Systematic Review Registrationhttps://www.crd.york.ac.uk/prospero/, identifier [CRD42020212253].