Twenty-one consecutive patients with displaced talar neck fractures (12 Hawkins type II, 9 Hawkins type III) were prospectively evaluated with magnetic resonance (MR) scans performed between 3 days and 12 months after surgery. All patients underwent open reduction and internal fixation with titanium screws, except two who underwent fixation with stainless steel implants that were subsequently removed. All patients had plain radiographs. We classified the scans as follows: type A, no abnormal signal changes in the body of the talus; type B, signal changes in less than 25% of the body; type C, signal changes in 25% to 50% of the body; and type D, signal changes in greater than 50% of the body. Plain anteroposterior radiographs correlated well with MR scans in patients with type D scans, but an inconsistent correlation was noted with lesser degrees of signal changes (types A–C), with the MR scans being more accurate in displaying the volume of avascular bone. Scans obtained less than 3 weeks after injury were not helpful in assessing for avascular necrosis. We found that high-quality MR images of the talus were consistently obtained in the presence of titanium screws in contrast to images obtained with stainless steel implants. We use titanium screws in all talar neck fracture repairs, because they permit high-quality MR images. We believe that further study of patients with Hawkins type III fractures and Hawkins type II fractures with equivocal radiographic evidence of avascular necrosis is warranted to try to identify those patients at risk for collapse and perhaps to guide weightbearing recommendations.