It is established that deep brain stimulation of the subthalamic nucleus improves motor function in advanced Parkinson's disease, but its effects on autonomic function remain to be elucidated. The present study was undertaken to investigate the effects of subthalamic deep brain stimulation on gastric emptying. A total of 16 patients with Parkinson's disease who underwent bilateral subthalamic deep brain stimulation were enrolled. Gastric emptying was expressed as the peak time of (13)CO(2) excretion (T(max)) in the (13)C-acetate breath test and was assessed in patients with and without administration of 100-150 mg levodopa/decarboxylase inhibitor before surgery, and with and without subthalamic deep brain stimulation at 3 months post-surgery. The pattern of (13)CO(2) excretion curve was analysed. To evaluate potential factors related to the effect of subthalamic deep brain stimulation on gastric emptying, we also examined the association between gastric emptying, clinical characteristics, the equivalent dose of levodopa and serum ghrelin levels. The peak time of (13)CO(2) excretion (T(max)) values for gastric emptying in patients without and with levodopa/decarboxylase inhibitor treatment were 45.6 ± 22.7 min and 42.5 ± 13.6 min, respectively (P = not significant), thus demonstrating levodopa resistance. The peak time of (13)CO(2) excretion (T(max)) values without and with subthalamic deep brain stimulation after surgery were 44.0 ± 17.5 min and 30.0 ± 12.5 min (P < 0.001), respectively, which showed that subthalamic deep brain stimulation was effective. Simultaneously, the pattern of the (13)CO(2) excretion curve was also significantly improved relative to surgery with no stimulation (P = 0.002), although the difference with and without levodopa/decarboxylase inhibitor was not significant. The difference in peak time of (13)CO(2) excretion (T(max)) values without levodopa/decarboxylase inhibitor before surgery and without levodopa/decarboxylase inhibitor and subthalamic deep brain stimulation after surgery was not significant, although motor dysfunction improved and the levodopa equivalent dose decreased after surgery. There was little association between changes in ghrelin levels (Δghrelin) and changes in T(max) values (ΔT(max)) in the subthalamic deep brain stimulation trial after surgery (r = -0.20), and no association between changes in other characteristics and ΔT(max) post-surgery in the subthalamic deep brain stimulation trial. These results showed that levodopa/decarboxylase inhibitor did not influence gastric emptying and that subthalamic deep brain stimulation can improve the dysfunction in patients with Parkinson's disease possibly by altering the neural system that controls gastrointestinal function after subthalamic deep brain stimulation. This is the first report to show the effectiveness of subthalamic deep brain stimulation on gastrointestinal dysfunction as a non-motor symptom in Parkinson's disease.