Introduction: The first generation of metal-on-metal bearings were introduced in Great Britain in 1960s. These bearings were abandoned already in early 1970s due to some problems and, primarily, due to a great success of Charnley's prosthesis. Since it was later realized that some of the metal-on-metal prostheses achieved almost 30-year survival, the interest in these bearings was renewed in 1990s. In this second generation the material was again Co-28Cr-6Mo; however, the alloy was wrought and not cast and can contain high or low carbon content. Characteristics: In vivo wear of metal-on-metal bearings is smaller (0.01 mm/year) compared to metal-on-polyethylene bearings (0.2 mm/year). Metal wear particles are usually nanometer-sized, from 25 to 50 nm, their number is approximately loo times larger than the number of metal particles formed on metal-on-polyethylene bearings. The effect of particles is twofold: local and systemic. Locally, changes in periprosthetic tissue related to a hypersensitivity-like reaction were noticed. Systemically, increased concentrations of metal ions in serum and urine were observed. Osteolysis was noted also in the second generation of metal-on-metal bearings. Clinical results: Some protheses from the first generation achived an up to 30-year long survival. The results for the second generation are currently in a mid-term phase. Depending on the type of material and design, differences in clinical performance are observed. Survivals between 80% at 8 years and up to 94% at to years were reported. Conclusions: Total hip replacements with metal-on-metal bearings show smaller wear and incidence of osteolysis compared to metal-on-polyethylene bearings. New problems may appear, however, related to a systemic increase of cobalt and chromium and a hypersensitivity to metals.