The microvascular anatomy of the proximal segments (M1 and M2) of the middle cerebral artery (MCA) was studied in 70 unfixed brain hemispheres from 35 cadavers. The arteries were injected with a tinted polyester resin and dissected under magnification by microsurgical techniques. The authors studied the outer diameter (OD), length, site of origin, and pattern of branching of the main trunk, secondary trunks, and the initial insular portion of the cortical branches of the MCA. The degree of mobilization of the arteries lying over the insular cortex was also assessed. The main trunk of the MCA, which had an OD of 3 +/- 0.1 mm bilaterally and a length of 15 +/- 1.1 mm in the right hemisphere and 15.7 +/- 1.3 mm in the left hemisphere, could be divided into four groups: Group I: absence of a main division (that is, a single-trunk type of MCA) (in 6% of cases); Group II: bifurcation (64%); Group III: trifurcation (29%); and Group IV: quadrifurcation (1%). The secondary trunks resulting from the division of the main trunk of the MCA had a mean OD ranging from 1.4 to 2.3 mm and a mean length that varied from 12.1 to 14.9 mm. The mean OD of the cortical branches measured near their origin in the main and secondary trunks indicated that the angular artery was the largest vessel, with a mean OD of 1.5 mm on both sides of the brain. The temporopolar artery was the smallest, with a mean OD of 0.8 mm in the right hemisphere and 0.9 mm in the left hemisphere. The authors also describe the patterns of origin of the cortical vessels from the main trunk (early branches) and from the secondary trunks, as well as their branching pattern at the site of origin (single vessels and common stems). These anatomical data indicate that it is possible to perform microvascular reconstructive procedures, such as anastomosis, grafting, and reimplantation of branches in the insular area. The advantages of using unfixed specimens, intravascular injections, and magnification to reproduce in vivo conditions as closely as possible are also discussed.