One hundred and seven patients were examined to compare advantages and pitfalls of the imaging methods currently used to stage parotid masses. Thirty patients underwent US and CT, 44 were investigated with US and MRI and 33 with US, CT and MRI. The accuracy of the three techniques was analyzed in detecting and assessing the intra-/extraglandular location of the parotid lesions and their benign/malignant nature. The intraglandular lesion spread to superficial/deep lobes was also investigated, together with its relationship to the surrounding structures. The findings were correlated with the cytohistological data from US-guided biopsy or from surgical resection. US was less accurate than CT and MRI, the latter being not statistically superior to the former. Because of the low incidence of pitfalls, US with fine-needle aspiration cytology should be the technique of choice; CT and MRI must be used for lesions > 3 cm or for masses in the deep gland lobes. In conclusion, CT accurately assesses parotid lesions but MRI demonstrates the relationship to adjacent structures better.