The sylvian fissure bifurcates posteriorly into ascending and descending rami. The diversity in the specific arrangement of these rami and in the length of a more anterior segment of the fissure, between the bifurcation point and Heschl's gyrus (segment H-B), was analyzed qualitatively and quantitatively, in both left and right hemispheres and in both males and females. Qualitatively, four basic patterns of bifurcation appeared: (A) the ascending ramus is larger than the descending one (61.25% of the cases). In half of these cases segment H-B was very short (designated as ‘short H-B’). The other patterns were: (B) the descending ramus is larger than the ascending one (7.5% of the total); (C) both rami are of approximately equal size (10% of the cases); and (D) both are of approximately equal size but the ascending ramus is oriented frontally instead of caudally as in the other cases (21.25% of the cases). Type D has not been reported before, and may be considered a new variant of bifurcating rami that in many cases corresponds to what other authors have referred to as the absence of an ascending ramus. We found a biased distribution of the fissure types according to hemispheres and also sex, with type A being more common in males and in the right hemisphere, and type D more common in females and in the left hemisphere. When the two hemispheres of each subject were matched, no correspondence was observed between the fissurization pattern of one hemisphere and that of the other, indicating that fissurization develops independently in each hemisphere. Quantitative analyses confirmed these findings and showed some new relations between components of the sylvian fissure. For example, when pooling together all fissure types a negative correlation between segment H-B and the ascending ramus was observed in males but not in females. On average, segment H-B was larger on the left side while the ascending ramus was larger on the right, confirming previous reports. Since earlier studies indicate that the planum temporale is larger on the left side, we suggest that the latter usually corresponds to segment H-B. However, the ‘short H-B’ cases described above have an unusually long and deep ascending branch, indicating that the planum temporale may run into the latter in these cases. The present classification of fissure types therefore describes a new variant of fissurization patterns in the sylvian fissure, which is asymmetrically distributed across the hemispheres and is perhaps sexually dimorphic. Furthermore, our analysis of fissure morphology and asymmetry is of direct relevance to the definition and location of the planum temporale in the sylvian fossa. Finally, our quantitative analyses are amenable to the use of morphometric techniques in the study of variability in fissurization patterns.