The assessment of the progression of tooth surface loss has until recently been limited to either the application of subjective ranking scales or visual comparison of sequential study casts. The development of quantitative measuring techniques offers the potential of greater accuracy and sensitivity. As direct intra-oral measurement is problematical such approaches often utilize impressions of the teeth, recorded at different epochs, to construct replicas for mapping and comparison. This in vitro investigation sought to determine the reproducibility of such an approach taking into account the total process chain. Two inlay cavities (one large, one small) were prepared in the palatal aspect of a plastic maxillary central incisor and restored with two flush fitting inlays. A series of impressions of this tooth were recorded, using a special tray and an addition cured light bodied silicone impression material (President, Coltene, Switzerland), with (a) both inlays in (b) both inlays out (c) large inlay out and small inlay in (d) large inlay in and small inlay out - a total of 16 impressions. Electroconductive replicas were fabricated from these and mapped using a computer controlled probe. Each series simulated wear of the tooth. A surface matching and difference detection algorithm was then used to compare each series of replicas and calculate the proportion of the surface undergoing simulated wear by a direct comparison of (a) matched to (b) or, indirectly as the summation of the results of matches of (a) with (c) and (a) with (d). The mean proportion of the surface with wear calculated directly was 26.6% (s.d. = 0.6) and indirectly 26.1% (s.d. = 0.5). A one-way ANOVA revealed no significant difference (P > 0.05). It is concluded that determining wear by this method is highly reproducible.