1. A total of 564 metal components from 109 patients have been examined.

2. Corrosion was detected on 228 components.

3. Most implants were removed for reasons other than corrosion.

4. In at least twelve cases corrosion was the reason for removal of the implant.

5. No corrosion of cobalt-chromium-molybdenum implants has been detected by the methods described in this paper.

6. Corrosion of ferrous alloy implants occurs in the human body.

7. The most common site for corrosion is the junction between components of implants.

8. The most corrosion-resistant type of ferrous alloy for implant uses is 18/8/Mo steel, which falls within specification En58.J of the British Standards Institute, and specification 316/317 of the American Iron and Steel Institute.

9. About 13 per cent of components removed (and by no means only when something was amiss) showed "face" corrosion when all the components of the implant were made of 18/8/Mo steel.

10. Four per cent of 18/8/Mo components of implants showed corrosion at sites other than the interface between components.

11. On the basis of corrosion resistance it is preferable to use cobalt-chromium-molybdenum alloys for implants that are to be left in the body for an indefinite period.

12. The corrosion resistance of the high alloy steels examined (18/8, 18/8/Ti, 18/8/Mo) does not appear to be related to hardness.

13. The marking of components, either by punching or by electrolytic methods, has not predisposed to corrosion.

14. All hollow 18/8/Mo implants should have a clean and metallurgically satisfactory internal surface.

15. The figures in this report do not permit a full statistical survey of corrosion in surgical implants because it has not been possible to examine a satisfactorily random sample. Many patients cannot be followed up and others die with the implant still in the body.