This metal-on-metal (MOM) retrieval study of products from four vendors (Biomet, Depuy, Smith & Nephew, and Zimmer) demonstrated that “normal” wear patterns on femoral heads (28–54 mm) were generally circular and occupied approximately 50 % of the hemispherical ball area. The main-wear zone (MWZ) centroid-vector averaged 16° superior to the polar axis, indicating that the wear pattern was mainly supero-medially located. Such wear patterns appeared very similar to that described on femoral heads from available simulator data. In contrast to femoral heads, the “normal” wear patterns on retrieved cups were both more variable, generally larger with 91 % of MOM cases showing breakout wear averaging 206° around the cup rim. Thus, the MOM retrievals demonstrated femoral wear patterns similar to those produced in simulators, whereas wear patterns in retrieved cups were more extensive than those produced in simulator tests. There has been a major focus on “edge wear” of MOM cups but surprisingly little or no attention given to the collateral damage on the femoral heads. In this study, we focus on the new definition of “micro-grooves,” i.e., third-body wear scratches 40 to 100 μm wide and 5–40 mm long. The major finding was the persistent “adverse” wear patterns that were apparent as well-defined stripe damage on both femoral heads and acetabular cups. The second observation was that the micro-grooves damage occurred preferentially at three sites, two of which, basal and polar stripes, were clearly indicative of cup impingement against the femoral implants (made of titanium alloy). The combination of basal and polar stripes showed a 100% incidence in this retrieval study. Basal stripe damage originated at or near the base of the femoral head and generally continued in a radial direction toward the main-wear zone. The contiguous damage was the polar stripe inside the main-wear zone. Confirmation of impingement damage was the presence of 40–100 μm-wide tracks of metal contamination that contained elements of the alloy used in the femoral stems (Ti, Al, and V). It is to be noted that the basal and polar stripes had 100% occurrence in this failed population of MOM cases. It may be that the MOM cases continuing to function well do not follow such extreme wear patterns. The combination of basal and polar stripes was observed on all retrieved femoral heads regardless of diameter, indicating a very consistent in vivo damage pattern for MOM bearings. Stripe wear has been frequently described on retrieved ceramic bearings but not on retrieved MOM. Stripe wear was always present at sites of impingement between the cup rim and femoral neck. The dominant form of surface scratches were the “micro-grooves” that varied 1–30 μm deep and had smooth transitioning terminuses. It could, therefore, be hypothesized that the trigger mechanism for adverse wear of MOM bearings came from cup-to-neck impingement episodes that released CoCr particles of size 40–100 μm. The circulating CoCr particles contributed the dominant abrasive wear process (micro-grooves) until becoming finely pulverized with repeated articulations. Such particles would finally decompose down to nanometer size particulates that would number in the trillions. The result was a combination of adverse, abrasive wear ending in the often-described “self-polishing” effect of MOM bearings. Such consistent impingement-stripe and micro-groove evidence in 100% of these MOM retrievals has implications in the design of all THA bearings.