Although they are subject to some limitations, sensitization tests such as the guinea pig maximization test (GPMT) have for many years provided a valuable basis for the identification of skin sensitization potential. Thus they have been used widely by regulatory authorities, such as those in Europe, as a means to identify significant sensitization hazards associated with new chemicals. However, the standard of performance of guinea pig sensitization assays has been demonstrated to be widely variable. Consequently, the OECD sensitization test guideline (the de facto world standard) has been updated to incorporate recommendations for action whose aim is to achieve a minimum standard of test conduct. The principle is that a test laboratory should be able to demonstrate an acceptable level of response using a moderately sensitizing chemical. A list of 3 such chemicals is provided, hexyl cinnamic aldehyde, mercaptobenzothiazole and benzocaine. It is our experience that whilst good results can readily be obtained with the first 2 of these, benzocaine is much more difficult. Using both the GPMT and the local lymph node assay (LLNA), an OECD-recommended screening test, benzocaine has given highly variable results. A range of from 0% to 60% positive in the GPMT was found and, in most tests, benzocaine would not classify as a skin sensitizer according to EU criteria. In the LLNA, from a series of 12 tests conducted in 2 laboratories, only occasional positive results were obtained. Furthermore, these positive results were not reproducible. Reasons for this variability are discussed. However, the main conclusion must be that benzocaine does not represent a useful moderately sensitizing positive control.