Hominoids and lorines are assumed to possess greater shoulder mobility than other primates. This assumption is based on morphological characteristics of the shoulder, rather than on empirical data. However, recent studies have shown that the glenohumeral joint of hominoids is not more mobile than that of other primates (Chan LK. 2007. Glenohumeral mobility in primates. Folia Primatol (Basel) 78(1):1-18), and the thoracic shape of hominoids does not necessarily promote shoulder mobility (Chan LK. 2007. Scapular position in primates. Folia Primatol (Basel) 78(1):19-35). Moreover, lorines differ significantly from hominoids in both these features, thus challenging the assumption that both hominoids and lorines have greater shoulder mobility. The present study aims to test this assumption by collecting empirical data on shoulder mobility in 17 primate species. Passive arm circumduction (a combination of glenohumeral and pectoral girdle movement) was performed on sedated subjects (except humans), and the range measured on the video images of the circumduction. The motion differed among primate species mostly in the craniodorsal directions, the directions most relevant to the animal's ability to brachiate and slow climb. Hylobatids possessed the highest craniodorsal mobility among all primate species studied. However, nonhylobatid hominoids did not have greater craniodorsal mobility than arboreal quadrupedal monkeys, and lorines did not have greater craniodorsal mobility than arboreal quadrupedal prosimians. Nonhylobatid hominoids and lorines had similar craniodorsal mobility, but this was due to a longer clavicle, more dorsal scapula, and lower glenohumeral mobility in the former, and a shorter clavicle, less dorsal scapula, and greater glenohumeral mobility in the latter. This study provides evidence for the reexamination of the brachiation, slow climbing, and vertical climbing hypotheses.