In this review, we (i) describe the nomenclature of Broca's area and show how the circumscribed definition of Broca's area is disassociated from Broca's aphasia, (ii) describe in detail how the gross anatomy of Broca's area varies between people, and how the definitions vary between studies, (iii) attempt to reconcile the findings of structural asymmetry of Broca's area with the differences in methodological approaches, (iv) consider the functional significance of cytoarchitectonic definitions of Broca's area, and (v) critically elucidate the significance of circumscribed regions of cortex for language lateralisation and language development. Contrary to what has previously been reported in the literature, asymmetry of Broca's area has not been reproducibly demonstrated, particularly on a gross morphological level. This may be due to major inconsistencies in methodology (including different anatomical boundaries, measurement techniques and samples studied) or that the sulcal contours defining Broca's area are so naturally variable between people making a standard definition difficult. Cytoarchitectonic analyses more often than not report leftward asymmetry of some component of area 44 and/or area 45. If a structural asymmetry of Broca's area does exist, it is variable, which differs from that of the functional asymmetry of language, which is more consistent. One reason for this might be that the link between cellular architecture, connectivity and language function still remains to be elucidated. There is currently no convincing explanation to associate asymmetry of Broca's area with the lateralisation of language. (Contains 7 figures and 3 tables.)