Examining the stability of individual differences in behaviorally induced heart-rate reactivity in children, 34 boys having a mean age of 101.9 months were presented a cognitive task to perform "as rapidly and accurately as possible" at each of 2 experimental sessions scheduled 1 week apart, and recordings of heart rate were obtained during periods of rest and task performance on each occasion of testing. Task-related heart-rate reactivity across the 2 experimental sessions revealed substantial and highly reproducible individual differences, as evidenced (1) by subsets of subjects representing clearly differentiated groups of heart-rate reactors and non-reactors, and (2) by correlation of subjects' mean heart-rate responses between the first and second sessions. Individual differences in heart-rate reactivity were independent of subjects' task performance.