The nosology and aetiology of sporadic adult-onset ataxia are poorly understood. The aim of the present study was to answer the following questions: (i) How many sporadic ataxia patients have a genetic cause? (ii) How many sporadic ataxia patients suffer from multiple system atrophy (MSA)? (iii) Is there a specific association between sporadic ataxia and serum anti-glutamic acid decarboxylase (GAD) or antigliadin antibodies? and (iv) What are the clinical features of patients with unexplained sporadic ataxia? The study was performed in 112 patients who met the following inclusion criteria: (i) progressive ataxia; (ii) onset after 20 years; (iii) informative and negative family history (no similar disorders in first- and second-degree relatives; parents older than 50 years); and (iv) no established symptomatic cause. Thirty-two patients (29%) met the clinical criteria of possible (7%) or probable (22%) MSA. The Friedreich's ataxia mutation was found in five patients (4%), the spinocerebellar ataxia (SCA) 2 mutation in one (1%), the SCA3 mutation in two (2%) and the SCA6 mutation in seven (6%). The disease remained unexplained in 65 patients (58%). We did not detect anti-GAD antibodies in any of our patients. Antigliadin antibodies were present in 14 patients, 10 patients with unexplained ataxia (15%) and 4 patients with an established diagnosis (9%). Patients with unexplained sporadic ataxia had a median disease onset of 56.0 years. Decreased vibration sense (62%), decreased or absent ankle reflexes (40%), increased ankle reflexes (39%), dysphagia (38%) and extensor plantar responses and/or spasticity (34%) were the most frequent extracerebellar symptoms. Compared with MSA, disease progression was significantly slower.