In Experiment 1, homophone and repetition priming were evaluated in 3 implicit memory tests. Homophone priming was comparable in size to repetition priming in the naming task, while reduced, but significant, homophone priming was obtained in the lexical decision and perceptual identification tasks. In Experiment 2, pseudohomophones (e.g., brane) were utilized as the distractor foils at test rather than pronounceable nonwords (e.g., terl). The homophone priming obtained in the lexical decision task in Experiment 1 was eliminated suggesting that phonological priming can be strategic. The role of orthography in homophone priming in the naming task was evaluated in Experiment 3. Homophone priming appears to depend upon the amount of orthographic overlap; orthographically dissimilar homophone pairs (e.g., bomb, balm) exhibited less homophone priming compared to similar pairs (e.g., bell, belle).