Increasingly, recent research has identified relatively high rates of autistic types of symptoms in a variety of genetic conditions, such as fragile X (Turk and Graham, 1997), tuberous sclerosis (Bolton and Griffiths, 1997), Angelman syndrome (Trillingsgaard and Ostergaard, this issue) and others (see Gillberg and Coleman, 2000). Detailed information about genetic conditions that have a higher than expected association with autism is likely to play at least some role in the search for genetic markers. However, such research is time consuming and costly, and before conducting comprehensive investigations on conditions that might have some association with autism, some means of determining whether a particular disorder has more than a chance relationship is required. In the present study the SCQ was used as a screening instrument to detect the possible presence of autistic symptoms in individuals with Cohen syndrome in order to determine whether a more detailed investigation of autistic symptomatology was warranted. Although eight families had at one stage been given a diagnosis of autism for their child, and the Cohen Syndrome Support Group was also aware of the findings of the earlier pilot study, indicating high rates of autistic symptomatology (Howlin, 2001), most parents in the study had neither sought nor been given a diagnosis of autism. They were thus far more na?ve than the parents involved in the study of Berument et al. (1999) who had already been interviewed on the ADI?R before completing the SCQ. Nevertheless, those individuals who scored at or above the PDD criterion on the SCQ were in every case confirmed as meeting criteria for an autistic spectrum disorder when they were subsequently assessed on the ADOS and ADI?R. The association was equally robust for individuals with moderate to severe learning difficulties and for those whose IQ scores fell within the mild to low average range. These findings suggest that the SCQ could well prove of value in ntal disorders, in order to determine whether further diagnostic assessment of autistic-type symptoms is warranted. It can be used as a postal questionnaire and is easy to complete. For example, in the initial stages of the project, when it was sent to the families of 76 individuals all but six returned it, fully completed. The high rate of false negatives (individuals who do not meet criteria on the SCQ but subsequently meet criteria for autism/autistic spectrum disorders on the ADI?R or ADOS) is clearly a problem, but it should be noted that in three of the six cases in which disagreement occurred, there were also discrepancies between the ADOS and ADI?R ratings. It may be that the sensitivity of the SCQ is reduced when used with individuals in whom the presence of autistic symptoms is more equivocal. Nevertheless, the findings indicate that, at least within this particular clinical group, if an individual met cutoff for PDD on the SCQ, that individual also met criteria for autism/autistic spectrum disorder on more detailed and well-validated interview and observational measures. Clearly, sample size was small and Cohen syndrome is a rare disorder, so that the findings may not be upheld in groups with other genetic conditions. However, the data certainly indicate the potential value of this instrument for identifying autistic spectrum disorders associated with other genetic or developmental conditions, or for comparing rates of autistic features between different conditions.