Individual subjects were allowed to learn at their own rates by adding more items on each trial, contingent on current recall, while reminding them of any items not recalled to assure learning of all items presented. The constantly increasing recall in such ongoing learning was sustained by remembering more items as part of larger recall units.