By 18 months of age, infants can link a novel word with the target of a speaker's gaze, suggesting that they are sensitive to the speakers' referential intentions. Adopting a procedure developed with human speakers, infants' ability to follow and use a nonhuman agent's gaze when learning new words was examined. A programmable robot acted as the speaker (Experiment 1). Infants followed its gaze toward the word referent whether or not it coincided with their own focus of attention but failed to learn a new word in either case. Infants correctly mapped words in both cases when the speaker was human (Experiment 2). While having eyes appears sufficient to elicit gaze-following in 18-month-olds, it does not suffice for the attribution of referential intentions.