Every brain is different.Predicting an individuals attention is a formidable challenge for neuroimaging.Recent evidence from network neuroscience suggests that whole-brain functional connectivity measured with fMRI contains information about individual differences in attention during task performance.Yet,direct evidence for relating task-free networks to the ensuing performance in attentional control is still lacking.With a simple model trained to learn a network topological mapping between task-free and task-driven activity,we show that brain networks at rest are significantly correlated with individual differences in attentional control performance.Remarkably,the same model can also predict symptoms of attention deficit hyperactivity disorder in children and adolescents from a completely independent sample.These results together provide direct evidence that resting-state networks are a fundamental mechanism for attentional control.