The reading skills of a girl with Williams Syndrome are assessed by a timed word-naming task. To test the efficiency of lexical and nonlexical reading, we considered four marker effects: Lexicality (better reading of words than nonwords), frequency (better reading of high than low frequency words), length (better reading of short than long words), and contextuality (better reading of words with one-to-one grapheme-to-phoneme mapping than words with context-sensitive graphemes). Results suggested that this girl has failed to develop efficient orthography-to-phonology mappings and is dependent upon a lexical-semantic reading procedure in which a word's pronunciation is activated after its meaning is obtained.