The suggested link between autism and cerebellar dysfunction formed the background for a Swedish clinical study in 2001. Thirty-two children (17 females, 15 males; mean age 12y, SD 3y 10mo; range 6 to 21y) with a clinical suspicion of non-progressive congenital ataxia were examined, and parents were interviewed about the presence of neuropsychiatric problems in the child. Twelve children had simple ataxia, eight had ataxic diplegia, and 12 had 'borderline' ataxia. All but one of the 32 children had a mild to moderate gross motor disability according to Gross Motor Function Classification System (15 were categorized as level I, 16 as level II, and one child as level IV). Neuroimaging and neuropsychological testing were achieved in most cases. There was a strong association between learning disability* and autism spectrum disorder (often combined with hyperactivity disorder) on the one hand, and both simple and borderline 'ataxia' on the other, but a weaker link between ataxic diplegia and neuropsychiatric disorders. A correlation between cerebellar macropathology on neuroimaging and neuropsychiatric disorders was not supported. Congenital ataxia might not be a clear-cut syndrome of cerebellar disease, but one of many signs of prenatal events or syndromes, leading to a complex neurodevelopmental disorder including autism and learning disability.