When we talk, we gesture spontaneously. Spontaneous gestures that accompany speech are regarded widely as playing an important role in communication. Iverson and Goldin-Meadow (1997, 1998, 2001) showed that early blind children and adolescents, who had never seen spontaneous gestures, gestured like sighted speakers. Conversely, Sasaki (1993) suggested that spontaneous gestures were rarely observed in early blind adults. In this study, the oral responses to nine tasks by early blind speakers who were born blind, or those who had lost their sight before three years of age, including “problem solving” and “explaining a concept” tasks, were videotaped and analyzed. The results indicated that early blind speakers gestured spontaneously, but that the gestures were not as distinct as they were in sighted speakers. There were individual variations in the frequency of spontaneous gestures, as in sighted speakers. The appearance of spontaneous gestures was related to the quality of the task and the rhythm of the gestures, and was qualitatively similar to that of sighted speakers. However, “iconic gestures” (McNeill, 1987) were rarely observed in early blind speakers, and the shape and movement of their gestures differed from those of sighted speakers. These findings suggest that spontaneous gestures appear without visual experiences, but that further development of spontaneous gestures varies considerably, and depends on an individual's visual experience.