BACKGROUND: Alcoholism is the most common cause of cerebellar dysfunction, yet estimates of the incidence of alcoholic cerebellar degeneration (ACD) vary greatly, with differences in methodologies contributing to these disparate findings. This study set out to characterize the frequency and pattern of clinical signs of ACD in an alcoholic group using the International Cooperative Ataxia Rating Scale (ICARS). METHODS: We compared the performance of 49 alcoholics and 29 control participants. The relative contributions of demographic and alcohol consumption variables to ICARS scores in the alcoholic group were also examined. RESULTS: The alcoholic group demonstrated significantly poorer performance on all of the ICARS subscales as compared with the control group. Within the alcoholic group, performance was more impaired on the speech scale than on all of the other scales, except the lower limb component of the kinetic scale, and less impaired on the oculomotor scale compared with all other scales. Years of heavy drinking and lifetime alcohol consumption correlated with total ICARS scores; however, maximum daily consumption was actually negatively correlated with ICARS scores. Of the alcohol history variables, years of heavy drinking was the best predictor of total ICARS scores, making a 19% unique contribution, followed by the period of abstinence from alcohol, which uniquely contributed 7% of the variance. There were high correlations between age and male gender and the alcohol consumption variables; however, age and gender were still found to uniquely contribute 5 and 7% respectively to the variance in total ICARS scores. CONCLUSIONS: ACD may affect up to two-thirds of chronic alcoholics. Assessing the number of years an individual has been drinking beyond a certain threshold can give a good indication of the likelihood of ACD. Age, gender, and the source of the clinical sample may significantly contribute to the prevalence of ACD and require further detailed investigation.