1. Currently available total replacement hip and knee prostheses were tested in a machine enabling flexion-extension movements to be applied whilst the prostheses were surrounded with Ringer's solution or other liquid and loaded within the physiological range.

2. Prostheses of which both components were made in cobalt-chromium-molybdenum alloy produced visible quantities of alloy particles, whose sizes ranged down to about 0·1 microns, and cobalt and molybdenum ions in solution.

3. No metallic or plastic particles were detected during tests on a hip prosthesis made of stainless steel and high density polyethylene.

4. The frictional moments in cobalt-chromium-molybdenum hip prostheses were higher than in stainless steel-polyethylene hip prostheses, by a factor of at least 2 to 1.

5. It is accepted that the conditions of these tests were probably more severe than in life, but the difference is held to be one of degree and not one of kind.

6. The particulate alloy debris, when injected in massive doses into the muscles of rats, gave an incidence of malignant tumours which was comparable to that already established for pure cobalt powder, whereas particles of several other metals, tested in the same way, gave no tumours.

7. It is argued that the particles which are known to be produced in at least some patients using cobalt-chromium-molybdenum total replacement joint prostheses constitute a risk of tumour formation which is certainly small, possibly negligible, but not accurately calculable at present.

8. The results of these tests, particularly the differences in frictional moment and in the production of particulate debris, suggest a preference for high density polyethylene as one component of a total joint replacement prosthesis.