Over the past 50 years very many different treatments have been promoted as bringing about significant improvements, or even cures, for children with autism. However, few interventions involve controlled studies of any kind; randomised control trials are virtually non-existent and when appropriate research methodology has been applied the results are generally far from positive. Recent research suggests that the most effective results stem from early intensive behavioural interventions. Although many questions remain concerning the optimal age at which treatment should begin, the intensity of treatment and the many other variables that may affect outcome, there is growing evidence of general strategies that can be effective in ameliorating the problems associated with autism.