Rotational CT studies have been previously used in whiplash-associated disorders (WAD) to document rotatory instability of the upper cervical spine thought to be due to alar ligamentous injury. More recently MR imaging has been employed to image such injury more directly. Our study aimed to assess the reliability and reproducibility of such MRI findings. In 12 WAD patients and six asymptomatic controls the alar ligaments were imaged in the coronal plane with an 0.5-T MRI system using a quadrature neck coil and applying a fast spin echo proton density/T2-weighted sequence (TR/TE/ETL 2,500/18 ms/16, FOV 140 mm, matrix 200 x 256, 16 x 3 mm slices, scan time 25 min). Images were graded for symmetry of imaging plane using a 3-point scale and also for presence of ligamentous injury with a 4-point scale, by two independent observers on two separate occasions. The alar ligaments could be identified in all cases. Asymmetry of the imaging plane was found to some degree in over half of the cases. Such images were much more likely to be graded as indicating injury. Of a total of 72 assessments, clearly and probably normal grades were given in 75%, and clearly or probably abnormal grades in 25%. Kappa values for intra- and inter-observer agreement were moderate to very poor, however, and the grading system could not reliably distinguish between patients and controls. It was concluded that with MRI techniques presently employed, alar ligamentous damage as a causative factor in WAD has not been proven.