OBJECTIVES: The development of dyskinesia in Parkinson’s disease (PD) patients may be associated with pulsatile stimulation of dopamine receptors, especially when using short-acting dopaminergic agents such as levodopa. Adjunctive therapy to levodopa with dopamine agonists may increase the risk of dyskinesia. This study quantitatively assessed the short-term incidence of dyskinesia in patients with advanced PD when treated with rotigotine transdermal patch (RTG) compared with ropinirole (ROP) and pramipexole (PPX) immediate-release (IR) and extended-release (ER). METHODS: A systematic review of randomized controlled comparative trials of RTG, ROP, PPX and placebo (PBO) on a background of levodopa therapy in advanced PD patients was performed, followed by a network meta-analysis. Trials published up to September 2011 were included. Incidence of dyskinesia was assigned as reported in each study. Study-level relative treatment effects were combined using both random-effect (base case) and fixed-effect models (sensitivity analysis) within a Bayesian framework using Markov Chain Monte Carlo methods in WinBUGS. RESULTS: Sisteen trials and 4446 patients were incorporated in the meta-analysis, which included 20 direct pair-wise comparisons. Patient characteristics were similar in terms of mean age at baseline, disease duration and severity (UPDRS III); however, average levodopa dosage differed across studies (range 272.9-843.4 mg/day). Under the base case assumption, odds of dyskinesia occurrence were comparable across dopamine agonists; odds numerically favored RTG over ROP-IR (OR: 0.70 [95% CI: 0.31, 1.59]), ROP-ER (0.33 [0.09, 1.15]) and PPX-IR (0.86 [0.49, 1.58]), but not PPX-ER (1.02 [0.42, 2.43]). The sensitivity analysis suggested that patients receiving RTG were estimated to experience fewer dyskinesias than those receiving ROP-ER (0.32 [0.11, 0.88]). CONCLUSIONS: This meta-analysis suggests that in levodopa-treated patients with advanced PD, dopamine agonists were comparable in terms of short-term odds of experiencing dyskinesia, with a numerical trend favoring RTG over ROP-ER. Further large-scale, confirmatory randomized clinical trials are required.