Objective: People with Down syndrome (DS) have higher variability in their motor skills compared to other counterparts without intellectual disability. Given that the effect of physical training on the variability and accuracy is unclear, the purpose of this study was to examine the effect of nine sessions of overhand throwing training on the variability and accuracy of overhand throwing in children with DS. Methods: Twenty-seven children with DS randomly assigned to experimental and control groups. In the pretest, two groups threw the tennis ball three times to the fixed target. Absolute error (AE) and the normalized root mean square (NoRMS) of segmental couplings of shoulder-elbow (NoRMS 1) and elbow-wrist (NoRMS 2) calculated from the scores of throwing and kinematics data, respectively. Then the experimental group participated in overhand throwing training for nine sessions. After that, the post-test was taken and ten days later, the retention test performed with pretest conditions. Results: The results of two-way ANOVA with repeated measures on AE values showed the experimental group in the post-test and retention phases was more accurate than the control group. Also, AE values of the experimental group in the post-test and retention phases were significantly lower than the pretest. The results of two-way ANOVA with repeated measures showed that NoRMS 1 in the experimental group was significantly lower than control group in the post-test and retention phases. Conclusion: It seems the overhand throwing training can reduce the variability and increase the accuracy of overhand throwing in children with DS.