Previous authors (Pugh and Gaze, 1966;Wagstaff et al, 1969) have reported favourably on the use of the Technicon AutoAnalyzer for automation of the CWR and RPCFT, but one of the main problems seemed to be the continuous flow system's limited rate of testing. If all specimens were tested against cardiolipin and Reiter antigens separately, and all those giving positive or doubtful results were retested against complement alone, it would be difficult to test more than 150 samples each day as allowance must be made for the time required to prepare the machine for use, to wash it through after testing is finished, and to transcribe the results. This would not be considered a significant improvement on the manual test. It is, of course, possible to