The current study examined the robustness, stability, reliability, and isolability of the attention network scores (alerting, orienting, and executive control) when young children experienced repeated administrations of the child version of the Attention Network Test (ANT; Rueda et al., 2004). Ten test sessions of the ANT were administered to 12 young children. Participants were asked to indicate the direction of a target fish, flanked by distractors, presented either above or below the fixation cross following different types of visual cues. Network scores, reflecting alerting, orienting, and executive control, were calculated using orthogonal subtractions of performance in selected conditions. Only the alerting network scores remained highly significant across the 10 sessions. The executive network scores showed some practice effects. The reliability of the network scores remained poor regardless of the amount of data. Based on the results, use of the ANT is cautioned against in young children when repeated testing is required.