The amnesic patient E.P. has demonstrated normal levels of repetition priming and at-chance recognition performance (S. B. Hamann & L. R. Squire, 1997), suggesting that the sense of familiarity used to make a recognition memory judgment is not based on the same mechanism responsible for repetition priming. However, the recognition tests previously used may have discouraged the use of familiarity and encouraged reliance on episodic memory. This issue was addressed in 5 experiments with E.P., 3 other amnesic patients with hippocampal damage, and 8 healthy controls. In Experiments 1-3, which were designed to discourage the use of episodic memory, the amnesic patients were impaired and E.P. performed at chance. In Experiments 4 and 5A, a stem-completion priming task was combined with a recognition memory task on each trial. E.P.'s priming was intact, yet his recognition memory performance was at chance. This suggests that although recognition memory judgments may be made on the basis of familiarity, repetition priming is not the source of this feeling of familiarity.