Although various drugs are currently used for restless legs syndrome (RLS) in clinic, selecting appropriate drugs for patients is difficult. This network meta-analysis (NMA) aimed to compare the efficacy and safety of different drugs. After literature searching and screening, 46 trials, including 10,674 participants are included in this NMA. The pooled results showed that, compared with placebo, only levodopa is inefficient to relieve symptoms of RLS. Cabergoline decreases IRLS scores to the greatest extent among all drugs (MD −11.98, 95% CI −16.19 to −7.78). Additionally, pramipexole is superior to ropinirole in alleviating symptoms of RLS (MD −2.52, 95% CI −4.69 to −0.35). Moreover, iron supplement alleviates RLS symptoms significantly compared with placebo in patient with iron deficiency (MD −5.15, 95% CI −8.99 to −1.31), but not for RLS patients with normal serum ferritin level (MD −2.22, 95% CI −6.99 to 2.56). For primary RLS, these drugs are also effective, while there is insufficient data to analyze drug efficacy in secondary RLS. We analyzed risk of common adverse effects of drugs including nausea, somnolence, fatigue, headache and nasopharyngitis. Alpha-2-delta ligands and DAs are favorable choices for both primary and secondary RLS because of their significant efficacy and good tolerability. Iron supplement can significantly alleviate symptoms of RLS patients with iron deficiency than placebo. We recommend gabapentin, gabapentin enacarbil, and pregabalin for clinicians for first consideration mainly because that they rarely cause augmentation. Oxycodone-naloxone could be considered in patients with severe or very severe RLS who failed in treatment with above drugs.