Objective: We screened a large series of individuals with congenital mirror movements (CMM) for mutations in the 2 identified causative genes, DCC and RAD51. Methods: We studied 6 familial and 20 simplex CMMcases. Each patient had a standardized neurologic assessment. Analysis of DCC and RAD51 coding regions included Sanger sequencing and a quantitative method allowing detection of micro rearrangements. We then compared the frequency of rare variants predicted to be pathogenic by either the PolyPhen-2 or the SIFT algorithm in our population and in the 4,300 controls of European origin on the Exome Variant Server. Results: We found 3 novel truncating mutations of DCC that segregate with CMM in 4 of the 6 families. Among the 20 simplex cases, we found one exonic deletion of DCC, one DCC mutation leading to a frameshift, = missense variants in DCC, and 2 missense variants in RAD51. All 7 missense variants were predicted to be pathogenic by one or both algorithms. Statistical analysis showed that the frequency of variants predicted to be deleterious was significantly different between patients and controls (p