Herberg and Montgomery (1987) recently described how behavioural factors can influence chlordiazepoxide (CDP) tolerance. Animals that received CDP before operant sessions showed tolerance to the sedative effect of CDP; the same treatment after operant sessions produced little tolerance. Such an effect is typically termed" Contingent" tolerance as tolerance is contingent upon animals receiving the drug in association with behavioural testing. Many studies indicate that contingent tolerance develops to a number of drugs in various behavioural and in vitro bioassays (Wolgin 1988). Collectively, the literature shows that for no single drug class have the necessary or sufficient conditions been established for tolerance to be behaviourally contingent (Goudie 1988). For example, there is much evidence that stimulants can show contingent tolerance in a variety of behavioural tests (Demellweek and Goudie 1983a, b) but there are also studies (e.g. Finnegan et al 1982) which report the development of tolerance to stimulants which was not behaviouraUy contingent.Herberg and Montgomery's (1987) data appear to contrust with our recent report (Gfiffiths and Goudie 1987) that tolerance to the sedative action of the benzodiazepine midazolam does not show a contingency effect. However, as noted by Herberg and Montgomery (1987), there were a number of procedural differences between these studies.