Essentially pure, ∼1.5 μm thick, polycrystalline diamond films DC–PACVD deposited on polycrystalline α-SiC flats were friction and wear tested against similarly coated α-SiC pins. The oscillatory sliding tests were performed with a Knudsen cell-type, wide temperature range tribometer built into a scanning electron microscope. Experiments were completed in 1.33 × 10−3 Pa (1 × 10−5 Torr) and 13.3 Pa (1 × 10−1 Torr) Pair, at test flat temperatures cycled to 850°C and back to room temperature. The results are compared with similar tests previously completed with diamond versus bare α–SiC and diamond versus Si(100) sliding combinations. In the absence of in situ surface analytical capability in the SEM tribometer, the findings are interpreted based on relevant information collected from the literature. The data indicate that the friction of the respective, tangentially sheared interfaces depends on the generation and annihilation of dangling bonds. Desorption of hydrogen during heating and sliding under the electron beam, in vacuum, create unoccupied orbitals on the rubbing surfaces. These dangling bonds spin-pair with others donated from the counterface, leading to high friction. Resorption of adsorbates such as hydrogen (e.g., by cooling the tribosystem) lowers the interfacial adhesion and friction. At high temperatures, in vacuum, the interaction energy may also be reduced by surface reconstruction and graphitization. At high temperatures, in Pair, the coefficient of friction of diamond versus itself is substantially lower than that in vacuum. This reduction is most probably caused by the generation of oxidation products combined with the temperature-shear-oxygen-induced phase transformation of diamond to graphite. The wide temperature wear rates of pure, polycrystalline diamond, characteristic of our test procedure, varied from ∼4 × 1016 m3/N · m in 1.33 × 10−3 Pa vacuum to ∼1 × 10−15 m3/N · m in 13.3 Pair.