Cases of Guillain-Barré syndrome (GBS) associated with parenteral use of gangliosides have been reported in several European countries. To evaluate the hypothesis of association between ganglioside exposure and occurrence of GBS, a case-control study was conducted. GBS cases discharged during 1989 from public and private hospitals in three Italian provinces were identified: 42 GBS cases and 420 controls matched on age and gender were enrolled. Data of onset of symptoms of GBS was taken from clinical records. Exposure status of subjects was ascertained through the regional computerized drug prescription monitoring system. The odds ratio of association between ganglioside use, in the 30 days prior to onset of symptoms, and GBS was 9.1 (95% confidence interval 2.8-29.4). Although there are formidable difficulties in distinguishing prodromal therapy of GBS from drug causation, the association with ganglioside therapy is strong and supportive of the hypothesis of a role of ganglioside preparations in the occurrence of GBS.