AbstractRates of wear have been determined for several elastomer materials, using a razor‐blade abrading apparatus based on one described by Champ, Southern, and Thomas. Measurements have been carried out at different levels of frictional power input, corresponding to different severities of wear, at both ambient temperature and at 100°C, and both in air and in an inert atmosphere. It is concluded that wear occurs as a result of two processes: local mechanical rupture (tearing) and general decomposition of the molecular network to a low‐molecular‐weight material (smearing). Marked differences were shown by different elastomers. Carbon‐black‐filled natural rubber, SBR (styrene–butadiene copolymer) and EPR (ethylene–propylene copolymer) were particularly susceptible to decomposition and smearing, but for natural rubber and SBR the decomposition process was not observed in an inert atmosphere. It is attributed to molecular rupture under frictional forces followed by stabilization of the newly formed polymeric radicals by reaction with oxygen, if present, or with other polymer molecules, or with other macroradicals. Cis‐polybutadiene and trans‐polypentenamer did not appear to undergo smearing to a significant degree. The reactive radicals formed in these materials by molecular rupture are assumed to undergo rapid addition to other molecules so that the network structure is maintained. Rates of wear have been found to increase with the applied frictional force raised to a power n. The value of n was between 2.5 and 3.5 for unfilled materials at ambient temperature, in agreement with Champ, Southern, and Thomas, who pointed out a general correlation with mechanical fatigue. Filled materials were found to be less sensitive to the frictional force, whether wear took place by tearing or by smearing, having values of the index n of 1.5–1.8. Several observations suggest that wear, even in the absence of smearing, is not fully correlated with mechanical fatigue: the markedly lower wear rates for carbon‐black‐filled materials, the anomalous rankings of unfilled materials, and the relatively small effects of raising the test temperature to 100°C. It is concluded that abrasive wear by small‐scale tearing is not accounted for solely by the crack growth properties of the material but involves other failure processes as well.