The aim of this study was to assess the efficacy and the safety of ondansetron administered orally in patients with a cerebellar disorder. The study was a randomised, multi-center, double-blind trial. The patients were randomised either to oral ondansetron 8 mg or to placebo twice daily for seven days. Cerebellar dysfunction was quantified before and after treatment using the International Cooperative Ataxia Rating Scale (ICARS). We performed a global analysis (total scores), we analysed by subscores (4 subscores: oculomotor, speech, kinetic, postural) and subgroups (4 subgroups: Cerebellar Cortical Atrophy (CCA), Multiple Systemic Atrophy (MSA), Familial Cerebellar Degeneration (FCD) and miscellaneous cerebellar disorders), and we also performed an analysis by individual test items. We investigated whether ondansetron and placebo had different effects upon ICARS total scores and subscores in the 4 subgroups considered together or separately. For p values < 0.05, we subsequently applied the Mann-Whitney test to compare ondansetron and placebo effect for each individual item. We evaluated 45 of the 46 patients included. No effect was found in global analysis. We found no difference in the analysis of the ICARS subscores. Concerning the individual test items, there was a significant difference between the placebo and ondansetron for the finger-to-nose test (p = 0.049), the Heel-to-Knee test (HK); (p = 0.03), the Body Sway Eyes Closed (p = 0.017) and the Body Sway Eyes Open (BSEO); (p = 0.014). There was no significant difference for tremor in upper limbs (p = 0.32) or for gait (p = 0.49). The Mann-Whitney test showed a greater effect of ondansetron than placebo for BSEO in miscellaneous disorders (p = 0.013) and for HK in FCD (p = 0.036), but ondansetron was deleterious for HK in CCA (p = 0.019). Our study showed no effect of oral ondansetron on global cerebellar dysfunction. The analysis by subgroups showed that the oral form of ondansetron (a) is deleterious for coordination in patients with CCA, (b) has no effect upon tremor in upper limbs, and (c) has a mild effect upon posture and coordination in lower limbs in some subgroups of ataxic diseases.