We reviewed a series of sixty-one revised total hip replacements in order to assess the accuracy of plain radiographs (all hips), arthrograms (thirty-one hips), and aspiration (sixty hips) in demonstrating the presence of loosening or infection. Plain radiographs correctly showed the status of the fixation of 92 per cent of the femoral components but of only 63 per cent of the sockets. They indicated loosening in only 37 per cent of the loose sockets. Arthrography improved the over-all accuracy in demonstrating the status of the fixation of the sockets to 80 per cent and increased the accurate detection of loose sockets to 89 per cent; however, it was no more useful than the plain radiographs in the assessment of the femoral components. Aspiration of the joint proved reliable in excluding the possibility of infection in fifty-nine of sixty hips. We concluded that femoral components are best evaluated for loosening by plain radiographs, but arthrography substantially improves the diagnostic accuracy for loosening of the socket. Aspiration of the joint accurately excluded sepsis as the cause of pain in all of the hips from which joint fluid was obtained.