We performed a placebo-controlled cross-over trial of riluzole (100 mg b.i.d.) in 10 patients with probable multiple system atrophy (MSA) administering riluzole and placebo for 4 weeks each with a 4-week washout period. Outcome measures evaluated short-term anti-Parkinsonian effects using the Unified Parkinson's Disease Rating Scale (UPDRS) subscales (UPDRS-II, activities of daily living; UPDRS-III, motor examination; sum of UPDRS-II and -III) before and at the end of each treatment phase. Delta values were calculated by subtracting the UPDRS scores measured at the end of each treatment arm from those before onset of each medication phase. Riluzole was generally well tolerated. There were no significant anti-Parkinsonian effects of riluzole comparing the UPDRS delta values for both treatment arms using the Wilcoxon signed-rank test. It is unlikely that riluzole treatment could have clinically meaningful anti-Parkinsonian effects in MSA. A trial assessing the disease-modifying potential of riluzole in MSA is underway.