The accepted approach to the interpretation of local lymph node assay (LLNA) data requires comparison of responses in the test groups with background activity found in concurrent vehicle-treated controls. However, of established value in the interpretation of toxicity test data is the use of historical control values that provide one criterion against which to judge the integrity of individual experiments. Specifically, the availability of robust and relevant historical control data permits examination of whether, in any individual experiment, control values fall within the expected range. With the most commonly used vehicle employed in the LLNA, acetone/olive oil (4 : 1) (v/v), the mean values, standard deviations and normal ranges are increasingly well established for a given laboratory, although there is some variation between laboratories, particularly with regard to expected ranges. Against this background, it is possible to identify (and, if appropriate, eliminate) a concurrent vehicle-control value that falls well outside the expected range. To explore critically the potential merits of this approach, one specific example is examined in detail.