BACKGROUND This study has two main goals: 1.) to determine the potential influence of dopaminergic drugs on sleep-disordered breathing (SDB) in Parkinson's disease (PD) and 2.) to elucidate whether NREM and REM sleep differentially impact SDB severity in PD. METHODS Retrospective clinical and polysomnographic study of 119 consecutive PD patients and comparison with age-, sex- and apnea-hypopnea-index-matched controls. RESULTS SDB was diagnosed in 57 PD patients (48%). Apnea-hypopnea index was significantly higher in PD patients with central SDB predominance (n = 7; 39.3±16.7/h) than obstructive SDB predominance (n = 50; 20.9±16.8/h; p = 0.003). All PD patients with central SDB predominance appeared to be treated with both levodopa and dopamine agonists, whereas only 56% of those with obstructive SDB predominance were on this combined treatment (p = 0.03). In the whole PD group with SDB (n = 57), we observed a significant decrease of apnea-hypopnea index from NREM to REM sleep (p = 0.02), while controls revealed the opposite tendency. However, only the PD subgroup with SDB and treatment with dopamine agonists showed this phenomenon, while those without dopamine agonists had a similar NREM/REM pattern as controls. CONCLUSIONS Our findings suggest an ambiguous impact of dopamine agonists on SDB. Medication with dopamine agonists seems to enhance the risk of central SDB predominance. Loss of normal muscle atonia may be responsible for decreased SDB severity during REM sleep in PD patients with dopamine agonists.