The authors report on a diagnostic pitfall which is often observed in magnetic resonance angiography (MRA) of intracranial vessels (time-of-flight 3D)--i.e., the absence of one of the two transverse sinuses. Fifty patients underwent MRA of the intracranial vessels with the TOF 3D (FISP) and the TOF 2D (FISP) techniques. Five patients were affected with sinus stenosis. MRA was performed with a 1.5-T superconductive magnet and a circular head coil. All patients underwent Magnetic Resonance Angiography with FISP 2D sequences (FA 40 degrees, TR 40 ms, TE 12 ms) acquired on coronal plane as well as FISP 3D (FA 15 degrees, TR 30 ms, TE 7 ms) acquired on axial plane. In the patients in whom FISP 2D sequences correctly demonstrated the whole confluence of sinuses, FISP 3D sequences always depicted the upper sagittal sinus, while the right transverse sinus was seen in 84.4% (38/45) of cases and the left transverse sinus in 24.2% (11/45) of cases. In the patients with a stenosis in the confluence of sinuses, FISP 3D sequences demonstrated two false-positive cases with respect to FISP 2D. The statistical analysis (McNemar test) of the results demonstrated the difference between 2D and 3D sequences in the detection of the two transverse sinuses to be statistically significant: p < 0.01 for the right transverse sinus and p < 0.001 for the left transverse sinus. In the study of the intracranial vessels (TOF 3D technique) the absence of one of the two transverse sinuses is suggestive of a diagnostic pitfall; the combination of 3D and 2D sequences answers this diagnostic question.