The Childhood Asperger Syndrome Test (CAST) is a 37-item parental self-completion questionnaire designed to screen for high-functioning autism spectrum conditions in epidemiological research. The CAST has previously demonstrated good accuracy for use as a screening test, with high sensitivity in studies with primary school aged children in mainstream schools. This study aimed to investigate test-retest reliability of the CAST in a high scoring sample. To this end, 73 parents filled in the second CAST (CAST-2) within approximately 2 months of the first administration of the CAST (CAST-1). Agreement above and below the cut-point of 15 was investigated. The kappa statistic for agreement (<15 versus > or =15) was 0.41. It was found that 70 percent (95% CI: 58, 80) of children did not move across the cut-point of 15. The correlation between the two test scores was 0.67 (Spearman's rho). The CAST shows moderate test-retest reliability in a high scoring sample, further evidence that it is a relatively robust screening tool for epidemiological research.