Accelerated wear tests involving small areas of contact are frequently used both as an aid to materials selection for dry bearings and for basic studies of wear. It is shown that whilst a general correlation may exist between the wear rates of polymer composites in accelerated tests and during operation as dry journal bearings, discrepancies can occur. The latter are primarily attributable to the differences in the apparent areas of contact. The various ways in which the apparent contact area may influence the wear of dry bearing materials—polymers and carbons—are examined and discussed. Particular attention is given to the role of transfer to, and polishing of the counterface metal during the early stages of sliding. The wear process of polymers and carbons during sliding, against relatively smooth metals is often assumed to be one of fatigue on a localised asperity scale. The implications of this process on the variations of wear rate with load and counterface surface roughness are examined. The use of very small contact areas, to increase the sensitivity of wear measurement, can give rise to misleading variations of wear rate with load, resulting from the trapping of debris within the contact area.