Behavioural studies can help to validate, modify and refine schemes for classifying receptors that are developed from electrophysiological and biochemical experiments. Drug discrimination constitutes one family of behavioural techniques that is being extensively used for studying subtypes of receptors, mainly because the methods often have remarkably high pharmacological specificity but can be applied to agents from a diverse range of classes. This article reviews briefly studies on agents acting through opioid and 5-hydroxytryptamine systems, where the results of the behavioural studies are very largely consistent with findings from other approaches. Many drugs used in such work have limited selectivity for putative subtypes of receptor, but little is known about how such compound pharmacological stimuli are processed in drug discrimination experiments. The characteristics of the discriminative stimuli produced by a mixture of drugs are discussed with respect to implications for effects of single drugs with multiple actions. Based on these initial experiments on discrimination of a mixture of nicotine and midazolam, it appears that the components of a compound pharmacological stimulus may be perceived and processed independently.