In two experiments, we examined whether hypermnesia occurs in cued recall when response bias is controlled by instructing subjects to generate more responses than they normally produce under standard cued recall instructions. Subjects processed 36 pairs of words using a relational processing, item-specific processing, or intentional learning strategy. A well-categorized list was presented in Experiment 1, whereas a loosely categorized list was used in Experiment 2. Three standard or forced cued recall tests were then administered. Hypermnesia was observed even when subjects were forced to guess. Furthermore, as in previous studies, relational processing resulted in greater net improvement than item-specific processing or intentional learning. We conclude that cued recall hypermnesia is a genuine phenomenon.