The potential of metallic copper as an intrinsically antibacterial material is gaining increasing attention in the face of growing antibiotics resistance of bacteria. However, the mechanism of the so-called "contact killing" of bacteria by copper surfaces is poorly understood and requires further investigation. In particular, the influences of bacteria-metal interaction, media composition, and copper surface chemistry on contact killing are not fully understood. In this study, copper oxide formation on copper during standard antimicrobial testing was measured in situ by spectroscopic ellipsometry. In parallel, contact killing under these conditions was assessed with bacteria in phosphate buffered saline (PBS) or Tris-Cl. For comparison, defined Cu2O and CuO layers were thermally generated and characterized by grazing incidence X-ray diffraction. The antibacterial properties of these copper oxides were tested under the conditions used above. Finally, copper ion release was recorded for both buffer systems by inductively coupled plasma atomic absorption spectroscopy, and exposed copper samples were analyzed for topographical surface alterations. It was found that there was a fairly even growth of CuO under wet plating conditions, reaching 4-10 nm in 300 min, but no measurable Cu2O was formed during this time. CuO was found to significantly inhibit contact killing, compared to pure copper. In contrast, thermally generated Cu2O was essentially as effective in contact killing as pure copper. Copper ion release from the different surfaces roughly correlated with their antibacterial efficacy and was highest for pure copper, followed by Cu2O and CuO. Tris-Cl induced a 10-50-fold faster copper ion release compared to PBS. Since the Cu2O that primarily forms on copper under ambient conditions is as active in contact killing as pure copper, antimicrobial objects will retain their antimicrobial properties even after oxide formation.