Inherited copper toxicosis in Bedlington terriers was 1st reported in 1975 and the entire Dutch population was examined from 1976 until the present for presence of the disease. To examine the effect on the prevalence of the disease of excluding affected dogs from breeding we have compared 2 time cohorts, the 1st consisting of dogs born from January 1, 1976, to January 1, 1986 (n = 155), and the 2nd of dogs born from January 1, 1990, to January 1, 1997 (n = 195). The diagnosis was made in the 1st cohort by evaluating liver biopsies, and in the 2nd cohort with a DNA marker. The population was also resolved into clusters of related dogs to analyze the familial distribution of the disease in the population and to search for ancient founders of the disease among the ancestors of sick dogs. Forty‐six percent of dogs examined between 1976 and 1986 had copper toxicosis. Eleven percent of dogs examined in the 2nd cohort had evidence of disease. This reduction was achieved while maintaining the already limited genetic heterogeneity of the population: the number of clusters and the mean relatedness between the clusters were similar in both time cohorts. The disease was evenly distributed over the clusters of related dogs in both cohorts. All ancestors had contributed to the distribution of copper toxicosis and no specific founders could be identified. This indicates that when the breed was established in The Netherlands, the disease was already highly prevalent in the founding dogs.