In conclusion, it would appear from present evidence that several statements can be made. First, MRI is the most accurate method of staging the periprostatic spread of carcinoma. Both the prostate and the regional nodes can be evaluated. The role of ultrasound is, at present, unclear. Second, the utility of CT has diminished with the emergence of high-quality MRI, but in situations in which MRI scans are inconclusive (e.g., in the assessment of paraaortic node disease, in which images are degraded by peristaltic movement), CT may be useful. Third, the choice of method for the assessment of nodal disease is a more difficult question. Although MRI and CT have replaced lymphography in most centers, the latter study, especially when combined with fine-needle aspiration biopsy, still is useful. Fourth, bone metastases are best assessed with scintigraphy, which can point to areas requiring supplemental plain radiography. An MRI study of the spine, although highly sensitive, is not practical. Fifth, MRI is superior to CT myelography in assessing spinal disease. Finally, the chest radiograph is the mainstay of assessing chest involvement.